Skip to main content

Full text of "The Philippine journal of science"

See other formats


£TOS 



THE PHILIPPINE 

JOURNAL OF SCIENCE 

ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 

Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 

EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D. ; J. R. WRIGHT, Ph. D. ; G. W. HEISE, M. S. 
J. C. WITT, Ph. D. ; T. DAR JUAN, A. B. ; A. H. WELLS, A. B. 

r. c. McGregor, a. b. ; h. e. kupfer, a. b. 

Volume XII 

1917 

With 11 Plates and 12 Text Figures 





MANILA 

BUREAU OF PRINTING 

1917 



CONTENTS 

No. 1, January, 1917 

Page. 

Brill, Harvey C. The fermentation of Philippine cacao 1 

Heise, George W. The interaction of chloride of lime with the 

normal constituents of natural waters and sewage 17 

Five text figures. 

Brill, HaRvey C. A chemical investigation of the seeds of Pangium 

edule and of Hydnocarpus alcalae 37 

Review 47 

No. 2, March, 1917 
Cox, ALVIN J. The study of copra and other coconut products 4!) 

Brill, Harvey C; Parker, Harrison 0.; and Yates, Harry S. 
Copra and coconut oil 55 

Parker, .Harrison 0., and Brill, Harvey C. Methods for the 

production of pure coconut oil 87 

Brill, Harvey C, and Parker, Harrison 0. The rancidity of 

Philippine coconut oil 95 

No. 3, May, 1917 

Wells, A. H. Destructive distillation of Philippine woods Ill 

Brill, Harvey C, and Alincastre, Cecilio. The possible maximum 

vitamine content of some Philippine vegetables 127 

Witt, J. C, and Reyes, F. D. The effect of calcium sulphate on 
cement 133 

Wright, J. R., and Heise, George W. The radioactivity of Philip- 
pine waters 145 

One plate and 2 text figures. 

No. 4, July, 1917 . 

Brill, H. C, and Wells, A. H. The physiological active constituents 

of certain Philippine medicinal plants: II 167 

Four plates. 

Brill, Harvey C. The antineuritic properties of the infusorial 

earth extract of the hydrolyzed extract of rice polishings 199 

Brill, Harvey C, and Williams, Robert R. The use of chaul- 

moogra oil as a specific for leprosy 207 

No. 5, September, 1917 

Brown, William H., and Arguelles, Angel S. The composition 
and moisture content of the soils in the types of vegetation at 

different elevations on Mount Maquiling 221 

Three plates and 1 text figure. 

Aguilar, R. H. A comparison of linseed oil and lumbang oils as 

paint vehicles 235 

One plate and 1 text figure. 

Heise, George W. The crater lake of Taal Volcano 247 

One plate and 1 text figure. 

iii 



iv Contents 

No. 6, November, 1917 

Page. 

Brill, Harvey C, and Thurlow, Leavitt W. Alcohol from discard 
molasses in the Philippine Islands 267 

Heise, George W. The radioactivity of the waters of the mountain- 
ous region of northern Luzon 293 

One plnte and 2 text figures. 

Heise, GEORGE W. The constancy in the radioactivity of certain 

Philippine waters 309 

Index 313 



J 



Vol. XII, Sec. A, No. 1 



January, 1917 



THE PHILIPPINE 

JOURNAL OF SCIENCE 



ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D.; J. R. WRIGHT, Ph. D.; G. W. HEISE, M. S. 

J. C. WITT, Ph. D.; T. DAR JUAN, A. B.; A. H. WELLS, A. B. 

R. C. MCGREGOR, A. B.; H. E. KUPFER, A. B. 




MANILA 

BUREAU OF PRINTING 

1917 



PUBLICATIONS FOR SALE BY THE BTTEEATJ OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



KTIiXOLOGY 

A VOCABULARY OF TEE IGOROT LAH- 

GTJAGE AS SPOKEN BY THE 

BONTOC IGOROTS 

By Walter Clayton Clapp 

Paper, 89 pages. $0.75. 



Order No. 40S. 

postpaid. 
The voo; 
•nd Enfllish-lgorot. 



In !gorol-Engli»h 



THE NABALOI DIALECT 
By 

and 
THE BATAXS OF PALAWAN 
By Edward Y. 

Order No. 403. Paper, $0.25; half mo- 

rocco, $0.75; postpaid. 
The Nabalol Dialect (65 pages, 29 
plates) and the Bataks of Pal 
pages, 6 plate*) are bound under one cover. 



HNOLOGY- Continued 



By Najeeb M. Salebby 

Order No. 405. Paper, 107 pages, 16 

plates, 5 diagrams, $0.25; half mo- 
rocco, $0.75; postpaid. 
This volume deals with the earliest 
written records of the Moros in Mindanao. 
i of the rulers of Maglndanao are 
in five folding diagram*. 



NEGRITOS OP ZAMBALES 

By William 

No. 402. Paper, 83 Pages, 62 

If morooco, $0.75; 

hs, many of which 

rhow orna- 

ng fire with bamboo, 

and various type* 



THE BATAN DIAEECT A3 A MEMBER 

OF THE I-rai.IP.PIHE GEOtTP . 

OP LANGUAGES 

By Otto Scheerer 
and 



INDUSTRIES 

PHILIPPINE HATS 
By C. B. Robinson 



This 
history ai 
In the PI 



THE BTJEANUB 



weeks, respe 
The 29 i 



By Herbert S. Walke 



bs. 1 


Order No. 412. 

plates, 1 m 


north- 


Considered fro 
tlcal utility. Mr. 



one of the m 
d by the Bun 

not a mere cc 
.s in the field s 
ns of which 



THE HISTC 
By Najebi 



By Charles S. Banks 



Moros 



In A Manual 

ried the 
ctual work with 




THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 



VOL. XII 



JANUARY, 1917 



No. 1 



THE FERMENTATION OF PHILIPPINE CACAO J 

By Harvey C Brill 

(From the Laboratory of Organic Chemistry, Bureau of Science, 

Manila) 

In an article by me 2 the statement was made that "the necessity 
for fermenting or sweating cacao is now generally acknowl- 
edged." This assertion "challenges trouble," declares the editor 
of Tropical Life, 3 since "no two experts seem agreed on this 
matter." However, the consensus of opinion appears to be 
with the above statement.* 

Booth and Knapp, of Messrs. Cadbury Bros. Ltd., state: 

In general, we believe that if the planter only allows ripe pods to be 
gathered, ferments for a reasonable period, cures with care, and keeps the 
beans dry they will have the right appearance, and that he will be producing 
the best that the types of trees on his plantation will produce. * * * 
We understand that unfermented cacao finds purchasers, but fermented 
cacao, always obtains the higher price ; unfermented beans are more difficult 
to shell, and they produce an inferior cocoa. Partially fermented beans 
suffer from the same defects. 

W. H. Johnson, F. L. S., director of agriculture, Southern 
Provinces, Nigeria, says : 

Fermentation is more generally practiced than hitherto, but the period 
of fermenting and curing is too restricted. 

S. H. Davies, of Messrs. Rowntree & Co., while insisting 
that the fermentation is due to the action of wild yeasts in the 
beginning and that the later action is due to true yeasts, believes 

1 Received for publication November 17, 1916. 
= The enzymes of cacao, This Journal, Sec. A (1915), 10, 123. 
'Smith, Harold Hamel, Trop. Life (1916), 12, 5. 

4 Booth, H. P., and Knapp, A. W., Proc. Third Internat. Cong. Trop. Agr. 
(1914), 225 et seq. 

146844 



2 The Philippine Journal of Science 1917 

that fermentation alone brings out the best flavor. Bainbridge 
and Davies 5 have shown that this flavor is due to the presence 
of an essential oil, which they believe is formed during the 
process of fermentation. 

The only discordant voice raised at this Congress was that 
of Professor Perrot, of the Ecole Superieure de Pharmacie, 
Paris, who reported an experiment in which he submitted 200 
kilograms of cacao, sterilized at the Ivory Coast, to one of the 
French chocolate concerns. When roasted, this cacao became 
fragrant and in no respect was inferior to the products obtained 
by fermentation in the same region. He states that the pulp 
was removed by means of potassium carbonate solution and that 
the color was a fine violet. These two properties, the tenacity 
of the pulp and the violet color of the ribs, are characteristic 
of the unfermented cacao and are reasons for fermenting, since 
both are undesirable. 

Knapp, 6 in discussing the Perrot method, states that the 
beans had a compact, cheesy interior, that they dried more slowly 
than the fermented beans, and that the process would be more 
costly than the present methods and would require skilled labor. 
He concludes with a plea for the encouragement of the use of 
the best-known methods of fermenting by the planters. In his 
book on cacao, van Hall 7 says of fermentation : 

All based on the same principle and have the same effect. This effect 
is the development of an essential oil, which gives the cocoa its peculiar 
aroma ; the conversion of part of the bitter-tasting compound, so as to lessen 
the bitter taste ; and, finally, the liberation of the theobromine, the substance 
which gives cocoa its peculiar tonic and stimulating properties. 

These prominent authorities agree, with the exception of 
Professor Perrot, that the fermentation of cacao produces an 
improved product even though the changes taking place are not 
completely understood. To obtain various data that might 
throw some light on this process, the experiments recorded in this 
paper were performed. However, before this phase of the 
work is taken up, several other points will be discussed. 

Recently I have had an inquiry from the Hershey Chocolate 
Company in which a method for the improvement of the color 
in poorly fermented cacao was requested. 

'Bainbridge, J. S., and Davies, S. H., Journ. Chem. Soc. London (1912), 
101, 2209. 

•Knapp, A. W., Trop. Life (1915), 9, 227. 

'Van Hall, C. J. J., Cacao. Macmillan & Co., London (1914), 201. 



xii, a, i Brill: Fermentation of Philippine Cacao 3 

In my article dealing with the enzymes of cacao the state- 
ment was made that no glucoside-splitting enzyme was found 
in the forastero cacao examined at that time. Since this article 
appeared, I have made an investigation of the enzymes of the 
criollo variety and found that an emulsinlike enzyme which splits 
amygdalin, setting free hydrocyanic acid, exists in the latter. 
This same enzyme occurs in the forastero type, though not in 
as great activity. The enzymes found in the criollo and in the 
forastero types are identical in character, but in general they 
exist in somewhat larger quantities or more active forms in the 
former than in the latter. None was found that was peculiar 
to either type, and for this reason the results of the investigation 
of the enzymes of criollo are not recorded. The main difference 
is one of intensity of activity. The list stands as summarized for 
the forastero type in the preceding paper with the addition of an 
emulsinlike enzyme that exists in the unfermented seeds in some- 
what greater activity than in the fermented product. The cor- 
rected list for the fermenting bean is casease, protease, oxidase, 
raffinase, diastase, invertase, and emulsinlike enzymes. 

The Philippine Bureau of Agriculture, through its inspectors, 
has made a census of the various districts of the Islands for 
the purpose of obtaining information regarding the quantity of 
cacao produced and the methods of handling it. In only a few 
provinces was there more than enough raised for local consump- 
tion, but in most of them the presence of trees was n< ted, thus 
demonstrating that cacao will grow in many places in the 
Philippines. No conscious effort is made to ferment the beans, 
and the methods of preparation are very crude. These methods 
consist in drying the beans, without preliminary treatment, in 
the sun from three to six days, rubbing between the hands with 
ashes or rice husks to remove the pulp previous to placing in 
the sun, or mixing with rice hulls and sand and treading with 
the feet and washing to remove the pulp and then drying in 
the sun. While the quantity grown at present is small, the 
fact that the regions in which cacao can be grown are wide- 
spread throughout the Archipelago is encouraging. The ex- 
perience of other cacao-growing countries lends hope to the 
belief that the Philippine Islands may become important as a 
cacao-growing country. Tables showing the production in other 
countries, as recorded by van Hall, 8 follow: 



The Philippine Journal of Science 



Table I. — Production of cacao by countries, in tons of 1,000 kilograms. 



Gold Coast 

Ecuador. 

San Thome .__ 

Brazil -- 

Trinidad- 

San Domingo- 
Venezuela 

Grenada 

Lagos 

German colo- 
nies 

Ceylon 

Fernando Po . . 

Jamaica 

Java.-- 

Surinam.- 

Haiti 

French colo- 
nies 

Cuba 

St. Lucia 

Belgian Congo. 

Dominica 

Columbia 

CoBta Rica 

Other coun- 
tries 



12, 946 
32. 119 



32, 956 
21,370 



2,378 
2,709 



20, 534 
31.564 
30,261 
33. 818 
23.390 
14,818 
16. 848 
5,441 
2,276 



3, 823 
3.570 
2,726 
3,216 
2,460 
1.897 
2.122 



Total 1193,753 206.337 219.562 243,819 230.000 



23,112 
36, 305 
36, 665 
29. 158 
26, 231 

17,251 
5,846 
2.978 



2,349 
1,743 | 

2.579 I 
2,043 



17,381 
5,948 

4,471 

4.404 
3.064 
3.000 
2,783 

2,460 



39. 500 
35, 500 
i 35, 500 



20,900 
12,500 
5.500 
3.500 

5,400 
3,500 
2,300 
3,400 



Increase or de- 
crease of 1912 
production ever, 
1908 production. 



, .eld crease of aver- 
f or 5 I age over 1908. 
years. 



Tons. 
26,554 
3,381 
6,772 
-2, 956 
-2, 470 



I 

i 

1,575 I 

1.412 I 



1.364 

1,251 
940 



P. cent, 
205.1 
10.5 



-28.3 | 
6.6 



97.2 
23.4 
-23.3 
26.2 
-15.fi 
-41.1 
-26. 2 



-35.6 
17.6 



33,231 
32, 177 



16,057 
5,579 
2,923 

4.165 

3,408 

2,675 

2,767 

2.375 j 

1,641 ; 

2,033 

1,446 ! 
1.485 : 



Tons. 
14, 346 
2.739 
4.603 

—779 



P. cent. 

110.9 

8.5 



52.1 
20. 2 

in. 8 

2.7 

-0.1 



79.6 
21.9 
23.0 
32.0 

-21.1 



700 j 87.5 | 1.160 j 360 , 45.0 



17.9 :218,868 25,567 11.6 



For the purpose of comparison and information the consump- 
tion of various countries is added. 
Table II. — Consumption of cacao by countries, in tons of 1,000 kilograms. 



Country. 


1908 


1909 


1910 


1911 


1912 




42, 615 
34,352 
20,445 
21,052 
15. 821 
5,821 
6,580 
18, 455 


53,379 
40.725 
23.254 
24, 264 
19,387 
6,684 
5.9S0 
21, 165 


50,315 
43. 941 
25. 068 
24.082 
19, 187 
9,089 
5,517 
23, 967 


58, 965 
50, 855 
27,340 
25,396 
23, 536 
9,852 
6,379 
27, 665 


57,000 
55.100 
26,900 
28. 100 
24.900 
10, 300 
5,300 
32.400 


















165,141 






229. S88 


240.000 | 









The total increase for 1912 in per cent consumption based 
on the 1908 consumption is 45.3, while the increase in produc- 
tion based on the 1908 yield is 17.9 per cent. 



xii, a. i Brill: Fermentation of Philippine Cacao 5 

The average increase in per cent consumption for the five 
years 1908-1912 based on the 1908 consumption is 24.9, while 
the increase in production calculated in the same manner is 
11.6 per cent. 

Comparing this increase in consumption, which in each case 
is practically double the increase in production, with the increase 
in production, it is at once apparent that the production is 
lagging behind. In other words, the present cacao-growing 
countries must increase their production more rapidly than 
they have been doing and new fields must be developed, or the 
rate of increase in consumption will decline. This disparity in 
increase of production and consumption cannot continue without 
an effect on the selling price. A rise in the selling price will 
lessen the sales and encourage interest in the cultivation. Rea- 
soning in this manner, the growing of cacao on a large scale 
would appear to be a profitable venture for the Philippine Islands 
to attempt. 

Very few localities in the Philippine Islands at the present time 
produce enough cacao for local consumption. Pampanga and 
Iloilo Provinces are the most important exceptions. The former 
reports the presence of five thousand cacao trees in the vicinity 
of Mexico alone. The cacao raised in this region is of good 
quality and appearance and compares favorably with that grown 
elsewhere. Two varieties are grown, criollo and a very fair 
quality of forastero. In Table III are given some data regard- 
ing samples of these varieties examined by the Bureau of Science. 

Table III. — Average weight of fruits and seeds of Philippine cacao. 



Type. 


I*' 
J= c 

£1 

3 « 


Weight of fruits. 


Weight of seeds. 


Weight of seeds -=- 
weight of fruit. 


Maxi- 
mum. 


Mini- 
mum. 


Aver- 
age. 


Maxi- 
mum. 


Mini- 
mum. 


Aver- 
age. 


Maxi- 
mum. 


Mini- 
mum. 


Aver- 
age. 


Forastero 


60 
60 


481 
631 


a. 
136 

190 


9- 
281 


<7- 

116 


(7. 

48 


0. 
69 


P. ct. 


P. ct. 
14.4 


P.ct. 
24.6 

' 





















The average weight of the fruit and the percentage of the 
seeds in the criollo is considerably greater than for the forastero. 
This is accounted for by the fact that more careful selection 
has been made of this variety for planting. The forastero is 
much more common and less difficult to obtain. The planting 
of criollo should be encouraged. 

The data in Table IV illustrate some fermentation experiments 
with these two varieties of cacao. 



The Philippine Journal of Science 



\\ ! I i ! 

I o ' ■ c : 

I a I i g | 

: & J J g ! 

i 1 1 I £ i 

- § 1 g s : 



■o * « +5 



a c a a c 



II 

41 J3 



33 n « m z «: « E3 &. ca z o ' 



fa &< 
5 S 2 



•OBOBO ,J ■>)■ <o <o ■»■ 

paiuaoijaj Xjp | * gj «S g ff 



3 

:N 1 



•0B0B3 13M -g 

uo paeeq »-sqiN | j^- 



1 3 O H S O 



S 3 



\8uiXjp 

pUB auI?U8UI 

I -jaj o^ snp Bsoq 



«> (O !0 C (D 16 ia 




xii. a. i Brill: Fermentation of Philippine Cacao 





d 6 6 




a 


§ « 
pi. fa 




D 






ft! 








fa 






i 








c 






fa 

c 

C3 






















■§ 






"8 








g 




« -i 
























£ <D 




















J2 




« !H 






° 










1 

c 






1 

i 


c 


5 5 | § § 






& i 








tl £ 


1 e -a ■§ o 






S 


W to fa fa £ 






to 


O 0> -* 00 00 C 


-r 


S3 


S " S " 


S3 


00 


CO "5 CO O r* CO 


CO 


5 


co co co to co 8 


CO 




g CO g o o g 


CO 










- 


s 


d ifl «' u 6 ^ 


s 


o 


00 O lO (N C- C- 


■» 


g 


lO ifi lO LS Id « 


s 


o 


IM O lO CO CO CO 


<o 




SS3SS33 




* 


" 


§ 


C- « CC t- fll 5 


1 


s 


1888888 


1 






*" 


to tt t- t- t- t 


■* 






s 










= 






tj -a -o -a -a o 












! ! ! ! ! o 


fa 


; 


! 1 . ! ! o 


o 










: i : i i s 









fa 




- s » J i s 


s 




ft. H O H <! g 


t- - 2 ■>• 1 


t~ 


<£> U> tj- tj- tj- £ 


6 > fa 


fa 


fa fa fa fa fa <j 


< 



8 The Philippine Journal of Science mi 

This cacao was fermented by placing the wet beans in beakers, 
at the top and bottom of which were layers of cotton to retain 
the juices given off when the beans underwent fermentation and 
to prevent the too rapid escape of the moisture from the beaker. 
Twice daily the beans were removed from the beakers and care- 
fully stirred in order that they might ferment uniformly and 
that they might come in contact with the air. They were kept 
in an incubator at a temperature of 37.5° C. This was neces- 
sary, since the small bulk of seeds allowed the heat from the 
fermentation to escape so rapidly that, if left in the open air, the 
fermentation would have been very incomplete. Under these 
conditions the highest temperature reached by any of the samples 
was 45° C. The criollo fermented much more rapidly than 
the forastero. Active fermentation was practically complete 
at the end of four days, and a considerably finer product re- 
sulted, judging by the odor, color, and other organoleptic prop- 
erties. More difficulty was experienced in fermenting forastero. 
The fermentation proceeded considerably more slowly, and the 
temperature did not rise high enough to kill the germ of the 
seeds in all cases ; consequently there was a tendency to sprout. 
In larger masses of the beans the temperature would undoubt- 
edly rise higher and the germ would be destroyed, but the above 
tendency illustrates the difference in the speed of fermentation 
of the two types under the conditions of the experiment. 

(a) Samples C-6-y and F-6-7 were sterilized by heating in 
hot water until the germ was killed and the activity of the 
enzymes destroyed. When the samples had become cool, yeast 
was added, and they were incubated. 

(6) The treatment of samples C-6-TD and F-6-TD was 
identical in all respects with that of samples C-6-y and F-6-7, 
except that taka-diastase was added instead of yeast. 

(c) Sample C-6-y-S had yeast added, but it was not sterilized. 

(d) Sample C-6-TD-S had taka-diastase added, but it was 
not sterilized. 

(e) Samples C-7-C and F-7-C had chloroform placed on the 
sterile cotton covering to prevent the introduction of yeast and 
of bacteria. From time to time additional amounts of chloro- 
form were added to take the place of the volatilized portion. 

(/) Samples C-7-T and F-7-T had toluene added instead of 
chloroform. The subsequent treatment was identical with that 
of samples C-7-C and F-7-C. 

(g) Samples C-7-A and F-7-A had alcohol added instead of 
chloroform or toluene. The subsequent treatment was identical 
with that indicated in paragraphs "e" and "f." 



xii. a, i Brill: Fermentation of Philippine Cacao 9 

Table V carries some further analytical data in regard to 
these samples of cacao. All these results are calculated on the 
basis of the water-free product in order that they may be 
directly comparable with each other and with the results listed 
in Table VI. All results were obtained from the shelled seeds. 
Table V. — Analytical data of Philippine cacao. 

[Numbers give percentages.] 



Sample No. 


(2 


i 

1 

J 

EH 


* 

°§ 

p 

.Q 
5 


I 


i 

in 


1 


1 

£ 

a 
c 

1 

< 


1 


1 

! 
I 


< 


3 

p 

1 

c 


*" c 

III 


C-U. 


48.13 


0.71 


12.13 


0.73 


2.08 


3.62 


4.83 


14.83 


16.25 


5.05 


2.39 


4.06 


0-2 


52.27 


1.27 


12.81 


0.63 


0.85 


5.14 


4.31 


14.13 


12.88 


5.57 


2.16 


3.39 


C-4 


52.06 


1.11 


12.25 


0.46 


0.11 


5.52 


4.27 


14.69 


13.13 


5.10 


2.43 


3.35 


0-6— 


51.00 


0.90 


12.06 


0.45 


0.00 


5.41 


4.04 


15.71 


11.88 


4.63 


2.30 


3.26 


C-6 


55.34 


1.10 


11.44 


0.10 


0.00 


5.85 


4.04 


15.43 


12.25 


4.92 


2.45 


3.13 


C-7 


53.13 


0.84 


12.19 


0.36 


0.00 


5.91 


3.65 


15.36 


11.25 


4.70 


1.80 


2.72 


C-6-y 


56.69 


0.84 


11.38 


0.27 


0.00 


4.58 


2.88 


11.12 


11.08 


2.95 


1.63 


2.77 


C-6-TD 


53.62 


0.75 


13.06 


0.49 


0.27 


4.81 


3.13 


11.14 


12.75 


2.93 


1.46 


2.35 


C-6-y-S 


50.79 


0.70 


12.69 


0.15 


0.00 


4.62 


3.55 


13.48 


11.38 


4.10 


2.55 


3.51 


C-6-TD-S 


50.11 


0.80 


13.43 


0.15 


0.45 


4.26 


4.11 


17.91 


12.83 


4.60 


2.60 


3.30 


C-7-C 


50.08 


1.06 


12.94 


0.39 


0.56 


3.96 


2.96 


14.26 


11.58 


5.12 


2.12 


2.70 


C-7-T.. 


46.12 


1.10 


13.19 


0.77 


0.22 


5.17 


3.27 


14.86 


13.50 


5.22 


2.58 


2.52 


C-7-A 


45.64 


1.20 


12.75 


0.92 


0.10 


5.36 


3.59 


14.63 


13.88 


4.90 


2.32 


3.82 


F-U 


43.67 


0.70 


12.88 


0.91 


1.71 


2.33 


6.20 


14.78 


17.13 


4.93 


2.42 


4.17 


F-2 


49.92 


0.80 


13.50 


0.13 


1.83 


6.00 


6.31 


14.61 


15.00 


4.77 


2.33 


3.67 


F-4 


49.93 


1.05 


13.00 


0.02 


1.38 


5.18 


4.83 


13.94 


15.00 


4.98 


2.42 


3.74 


F-5 


51.09 


0.85 


13.13 


0.25 


0.22 


5.42 


4.31 


14.76 


13.07 


4.55 


2.23 


3.02 


F-6_._. 


52.21 


0.88 


13.91 


0.21 


0.45 


5.49 


4.20 


14.21 


14.13 


4.68 


2.20 


2.82 


F-7 


53.55 


1.02 


13.00 


0.10 


0.27 


5.92 


3.69 


14.75 


12.13 


4.77 


2.57 


2.99 


F-6-y 


50.30 


0.75 


13.38 


0.24 


0.00 


4.52 


3.07 


10.65 


11.07 


2.55 


1.16 


2.16 


F-6-TD 


55.99 


1.00 


13.88 


0.05 


0.40 


5.07 


3.35 


10.71 


8.85 


3.25 


1.38 


2.03 


F-7-C 


46.29 


1.08 


13.50 


0.33 


0.64 


4.15 


5.09 


14.39 


14.85 


5.27 


2.12 


2.51 


F-7-T 


48.36 


0.80 


13.75 


0.44 


0.30 


4.18 


3.55 


14.88 


12.03 


5.00 


2.24 


2.91 


F-7-A 


41.67 


0.85 


13.69 


0.87 


0.61 


4.56 


3.55 


14.44 


12.50 


4.98 


2.22 


2.78 


Average of C-U 


























to C-7, inclu- 


























sive 


51.99 


0.97 


12.15 


0.46 


0.51 


5.24 


4.19 


14.69 


12.96 


4.99 


2.22 


3.32 


Average of F-U 


























to F-7. inclu- 


























sive 


50.06 


0.88 


13.24 


0.27 


0.98 


4.89 


4.76 


14.51 


14.41 


4.78 


2.36 


3.40 



» For general methods of analysis, see Parry, E. J., Foods and Drugs. Scott, Greenwood & 
Son, London (1911), 1, 20. 

For purposes of comparison some figures taken from a paper 
by Ridenour 9 are entered in Table VI. I have recalculated these 
to the dry weight of the cacao in order that they may be directly 
comparable with the results of Table V. 



"Ridenour, William E., Am. Journ. Pharm. (1895), 67, 207. 



10 



The Philippine Journal of Science 





■ A . 






1 




Averag 
for Phi 
ippine f 
rastero 


50.06 
0.88 

13.24 
0.27 
0.98 
4.89 








3 3 " 


1 




lie c 
go. -si 


»S»!S3S 


s s s 






-i d n id 


S £J ■* 
















<o-« 












Sg?S!35SSS? 






>5f 


OHOHHTJldlOOr 






bE 


"ScnSmSSoSa 






si 


g t- ~ M N C- <N 






x E 


c-N^oic-ococowtj: 














SS 








si 


t-a»4aio^j , a>otoooc 






>> 








ss 


•» FH ,-( 






— 


"oo"§S"noSS 














OT corn 1 1C « « 1 


1 ■ 




a 




| 




A 


gsnns^s; 






-§8 

H 






10 




1 




5SS55SSSSSS 




*J> 


s-s 


<£> O »-H CDlO-VOCs 


? 









■3 


1 


So 

2 § 


9.42 
1.20 
1.33 
2.94 

1.67 
4. 08 
3.51 
7.60 
3.61 


« 
















y 


■3 




>• 


c 


<i 


to O SO .-< -V ■«* O* a 1 


1 


c 

3 


'E-O 


sgg;ss§ssss 


£ 












H-° 


"* r " 1 rt 






d 


SSgSS!S3SSS 






d 










** * H 






•E £ 


S8SS3£s53Ss!=; 








m or-i ed t t- jj e- 






W C 






Dl 


C-rHC»i-llOOCQtOlOOl! 






a] 


^JrHt-rH CO 00 3 Ol P" 






















































« 








1 


1 Theobrom 
Albumino 
Glucose . 
Sucrose _ 
Starch __ 


c 
I 


0J .t 

1 s 

w < 





Brill: Fermentation of Philippine Cacao 



11 



In Table VI a comparison is made of the analytical data 
obtained for Philippine cacao and a number of foreign cacaos. 
This comparison places the former in a very favorable light in 
so far as such data are indicative of the quality of cacao. 

An examination of the cacao butter was made to determine 
what changes, if any, took place in it with the variation in 
fermentation. 

Table VII. — Properties of cacao butter from Philippine cacao. 



c-u. 

C-2._ 



0-6... 

C-7... 
C-6-y 



C-6-TD ... 
C-6-y-S... 

C-6-TD-S 
C-7-C .... 

C-7-T 

C-7-A 

F-U 



191.5 
191.1 



192 6 
192.4 
193. 9 

194.8 
191.0 

191.-! 
192.2 

193. 6 

191. 7 

192. 7 



Acid 
value i 

cc. Index 
0.1 iV of re- 
base ! frac- 

per tion. 

oil.' I 



F-2.... : 196.5 



F-6._. 
F-7.__ 
F-6-y 



F-6-TD 



0.10 
0.17 
0.27 
0.23 

0.44 
0.13 



F-7-C.... 

F-7-T 

F-7-A 

Average for C-U 
to C-7, inclu- 
sive. 

Average for F-U 
to F-7. inclu- 
sive. 



197.0 
191.6 
192. 6 



0.49 
0.18 
0.26 



1.4600 

1.4599 

1.4582 

1.4580 

1. 4579 
1. 4579 



1.4579 
1. 4581 



1. 4585 
1. 4582 

1. 4580 
1.4581 
1. 4582 



Light yellow. Deeper 
than straw yellow. 

Somewhat deeper than 
C-U. 

Same as C-2 



1. 4579 

1. 4581 

1. 4582 
1. 4586 



do 



do 

do 

Slightly lighter than 

C-6-TD. 

do 

Almost water white 

.....do 

Deeper yellow than C-7-C. 
Same as C-U. Similar to 

C-2. 
Somewhat deeper than 

F-U. 

do 

do 

Deeper yellow than F-5 .. 
do 

Slightly lighter than F-7 . 

Deeper yellow than F-6-y _ 

Same as F-6-y 

do 

do 



Pronounced; somewhat 

harsh. 
Similar to preceding. 

Better than C-2. Faintly 

harsh. 
Good; not so pronounced at 

C-4. 
Pleasant; esterlike. 

Do. 

Mild; not so characteristic 

as C-7. 

Do. 

Do. 

Do. 
Very mild; slightly sweet. 
Not bo mild as C-7-C 
Pleasant; chocolatelike. 
Pronounced; not so evi- 
dent as C-U. 
Similar to F-U. 

Slightly better than F-2. 

Good; better than preced- 
ing three. 

Esterlike; pleasant. 
Do. 

Faintly esterlike; less so 
than F-7. 

Slightly harsher than 
F-6-y. 

Not pronounced; pleasant. 

Very pleasant. 
Do. 



12 The Philippine Journal of Science 1917 

DISCUSSION 

The loss in weight due to fermentation is given in Table IV. 
The criollo has the greater percentage of loss from fermentation. 
This is more apparent because the fermentation of the criollo 
was more nearly complete than the fermentation of the foras- 
tero. If perfect fermentation could be induced in the latter, 
the loss in it would undoubtedly approach that for criollo. 
However, a difference in the selling price of the fermented pro- 
ducts would be made in favor of the criollo because of its 
superior quality, and this would still allow more profit to be 
made from the cultivation of the criollo than from that of the 
forastero. Then, too, the latter, because of its slowness in fer- 
menting, is in greater danger of being spoiled by molding or 
by sprouting than is the former, and losses from this cause 
would be more apt to occur. On the other hand, the criollo 
samples examined had a larger average weight for the fruit, 
531 grams, and a higher average yield of seeds, 27.9 per cent, 
than the forastero, which had an average weight of 481 grams 
and an average yield of 24.6 per cent of seeds. This would 
make the yields of dry, shelled seeds compare, criollo to foras- 
tero, as 1,089 to 1066; consequently, regardless of a difference 
in price in favor of criollo, the profit from its cultivation would 
be greater than from the cultivation of forastero, since the 
absolute yield of dry, shelled seeds, or nibs, is greater from the 
criollo than from the forastero. 

The greater rapidity of the fermentation of the criollo is 
demonstrated by the loss in weight due to the fermentation. 
The maximum loss in weight has been reached at the end of 
the fourth day, while the forastero shows a maximum at the 
end of the sixth day. The intensity of the fermentation is like- 
wise demonstrated by these same figures, for criollo shows a 
maximum of 8.5 per cent and forastero a maximum of 3.8 per 
cent. The changes effected by fermentation should be much less 
in the case of the latter, judged by the change in weight, and 
this is borne out by the less agreeable odor of the defatted 10 
cacao and the cacao butter. The odor is most pleasant between 
the third and fourth days for the cacao butter and about the 
fourth day for the defatted cacao in the case of the criollo type, 
while the fifth and sixth days for the cacao butter and the 

" By defatted cacao is here meant cacao from which all the fat has 
heen etherized. 



xii, a. i Brill: Fermentation of Philippine Cacao 13 

Qfth and sixth days for the defatted cacao show the finest 
flavored product in the case of the forastero. But the product 
of this longer fermentation period for forastero does not equal 
that from criollo in quality. 

The changes brought about by fermentation are hard to de- 
monstrate by an analysis of the finished cacao. The changes are 
largely characterized by an improvement in its organoleptic 
properties. Bainbridge and Davies " believe that an essential 
oil is produced during fermentation. In their investigation on 
fermented arriba from Ecuador, they obtained 24 grams of an 
essential oil from 2,000 kilograms of cacao. Ordinary analytical 
chemical methods would not suffice for the detection of so minute 
quantities. The theobromine shows no regular variation. 
Sack 12 claims that one of the results of fermentation is the 
splitting of a glucoside with the formation of theobromine and 
cacao-red. The results obtained in this investigation do not 
corroborate his conclusions. 

The glucose and sucrose content change with the degree of 
fermentation, the latter in several cases being zero, while the 
former first decreases and then shows a slight increase. This 
increase is due to the action of diastase on the starch. The 
sugar usually listed as glucose in the analysis of fermented cacao 
is doubtless largely maltose, the product of the action of dia- 
stase on starch. The percentage of starch does not undergo a de- 
cisive change. The apparently smaller amount in C-U and F-U 
is partly accounted for by the fact that these samples have not 
lost juices through fermentation. I believe a change has taken 
place in the character of the starch by the fermentation and that 
this makes itself apparent in the so-called break of fermented 
cacao. 

A change in the percentage of astringent matter with length of 
fermentation is the most apparent change that can be demon- 
strated by analytical data. The amount decreases with the 
length of fermentation and accounts for part of the improve- 
ment in the flavor of the product. The superiority of the 
criollo over the forastero is apparent when a comparison is made 
of the quantities of astringent matters present in the two. 

The difference in the cellulose content of the fermented and 
unfermented samples can be explained by the fact that the 
former has lost juices by fermentation while the latter still re- 
tains these. The same explanation holds for the difference in 
extractive matter. Van Hall 13 states that good cacao should 

u Loc. cit. "Bull Dept. Agr. Surinam, 10. " Loc. cit. 



14 The Philippine Journal of Science 1917 

show 12 per cent or more extractive matter. The Philippine 
product averages slightly more than this. 

The free acidity of the cacao butter increases with the dura- 
tion of the fermentation. Criollo is superior to forastero in 
this respect also. In spite of the greater intensity of the fer- 
mentation, it shows less free acidity than does forastero, the 
ratio being 0.26 cubic centimeter to 0.90 cubic centimeter 0.1 N 
alkali, respectively. The slight change in the acidity of the 
cacao butter with the greater length of fermentation is confirm- 
atory of the conclusion reached by me 14 that Philippine cacao 
does not contain a lipase. 

For the purpose of making a further investigation of the 
fermentation processes, certain special samples were prepared. 
These are described on page 8. An extended discussion of 
these samples is hardly necessary, since in no case is the quality 
of the final product equal to that obtained by the use of the 
ordinary methods. Samples C-6-y and F-6-y had yeast added 
to them after they had first been sterilized to destroy the activity 
of the enzymes. In the sterilization of the seeds care was taken 
to prevent cooking and at the same time effectively to sterilize. 
The final product did not equal C-6 or F-6 in excellence of odor, 
color, or break. It would appear that yeasts alone do not pro- 
duce the desired changes. Bainbridge and Davies 13 make the 
statement that wild yeasts from the pods and stems first act 
on the pulp and that in the latter stages of fermentation their 
place is taken by true yeasts. The terms wild yeast and true 
yeast as used by these authorities are rather confusing. I do 
not believe the yeasts alone account for the changes cacao under- 
goes when it ferments, and the above results corroborate my 
belief. The yeast used for inoculation was the yeast found 
growing on the cacao; consequently no error could have arisen 
from a choice of the wrong yeast. 

For purposes of comparison and control, C-6-TD, C-6-TD-S, 
C-6-y-S, F-6-TD, F-6-TD-S, and F-6-y-S were run. These 
were no better than, and in some cases not so good as, C-6-y and 
F-6-y; therefore they add weight to the above conclusion that 
yeast is not solely responsible for the fermentation changes of 
cacao. 

To determine, if possible, the influence of the enzymes on the 
cacao, samples C-7-C, C-7-T, C-7-A, F-7-C, F-7-T, and 
F-7-A were prepared. Extreme care was taken to prevent 
inoculation of the seeds when the pods were opened and the 

" Loc. cit. " Loc. cit. 



joi, a. i Brill: Fermentation of Philippine Cacao 15 

seeds removed. In this endeavor I was successful, since at no 
time was any alcoholic or acetic acid fermentation apparent. 
These samples were examined in the same manner as the pre- 
ceding. The quality of the cacao was better than that produced 
by the action of yeast and taka-diastase in the absence of the 
enzymes, but the odor was not so fine as C-6 in the case of the 
criollo samples so treated, nor so fine as F-6 in the case of the 
forastero samples. Judging by these results, it would appear 
that the enzymes already existing in the cacao alone do not 
bring about all the desired changes, but that their influence 
must be reenforced by the enzymes from yeasts and possibly 
from bacteria or molds. The superiority or the peculiarity of 
certain cacaos is probably largely due to the presence of certain 
yeasts or molds during the fermentation process. It is instructive 
to cite the belief of Preyer 16 in this regard. He has isolated a 
yeast, Saccharomyces theobromae, from fermenting cacao. He 
recommends its use in the initiation of the fermenting process. 
The use of a pure culture yeast would necessitate extreme care 
in handling to prevent inoculation of the cacao with wild yeasts 
until the yeasts used for inoculation had attained a good growth. 
It would be impossible to sterilize the seeds, since such treat- 
ment would likewise destroy the enzymes, and the resulting 
product from enzyme-free cacao would not be satisfactory. The 
seeds as they exist in the pod are free from yeasts and bacteria, 
but handling of this in such a manner as to prevent contamina- 
tion is hardly practical. However, care in fermentation will 
yield a good grade of material provided the initial product is 
high class, so there is no cause for disconsolation because of 
the apparent impracticability of using pure cultures of yeast. 

SUMMARY 

Philippine cacao is compared with foreign cacaos. 

A study is made of criollo and forastero cacao fermented 
under varying lengths of time, and the respective influence of the 
enzymes and of yeast is investigated. 

The conclusion is reached that the Philippine Islands can 
grow a good quality of cacao in large quantities and that the time 
seems opportune for such an innovation. The investigation 
leads to the belief that the fermentation is the joint result of 
the reaction of yeasts and of enzymes. 

"Preyer, A., Tropenpflanzer (1901), 5, 157. 



THE INTERACTION OF CHLORIDE OF LIME WITH THE NORMAL 
CONSTITUENTS OF NATURAL WATERS AND SEWAGE " 

By George W. Heise 

FIVE TEXT FIGURES 

(From the Laboratory of General, Inorganic, and Physical Chemistry, 
Bureau of Science, Manila) 

In the course of an extensive study of the sterilization of 
water and sewage, parts of which have already been reported, (9) 
it became necessary to do considerable work on the decomposi- 
tion of chloride of lime (calcium hypochlorite) in water and 
on its interaction with waters and sewage. The work was done 
because of its practical importance, and no attempt was made 
to study in detail the chemical reactions involved ; the results are 
presented at this time because of their possible application to 
sterilization problems involving the use of hypochlorite solutions. 

The value of chloride of lime as a disinfectant was first pointed 
out by Koch (12) in 1881, while its application on a large scale to 
the sterilization of water for municipal supply was first proposed 
in 1894 by Traube. (23) That waters containing hydrogen 
sulphide or relatively large quantities of organic matter are not 
readily sterilized with chloride of lime was shown by Lode (14) 
in 1895. Since that time hypochlorites have been so widely used 
for the disinfection of water and sewage that their decomposition 
and their interaction with the substances normally found in water 
and sewage have been much studied. 

Hypochlorites in distilled-water solution decompose, even in 
the dark, with measurable velocity. (2) Both chlorate and 
oxygen are formed, although the main reaction proceeds accord- 
ing to the equation 

2NaC10 -> 2NaCl + 2 . 

The decomposition is accelerated by heat, proceeding in 
accordance with the equation : (2) 

3NaC10 -> NaCIO, + 2NaCl, and 
2NaC10 -> 2NaCl + 2 . 

The results of Bhaduri(2) indicate that for certain concen- 
trations in the dark at 100° C. the reaction is monomolecular, 

1 Received for publication December 1, 1916. 



18 The Philippine Journal of Science im 

although its order has not been determined experimentally. The 
photolysis has been studied quantitatively by Lewis, (13) who 
concluded that the reaction was probably monomolecular. 

When waters contain foreign substances either in solution or 
in suspension, the decomposition rate of hypochlorites is gen- 
erally markedly changed. Bhaduri(2) found that in the dark a 
sodium hypochlorite solution was most stable when the con- 
centration of sodium hypochlorite was 1.5-1.7 per cent; of salt 
molecules, approximately 0.4 per cent. In the light, (13) alka- 
line is more stable than neutral sodium hypochlorite solution. 
Concentration and temperature have very little effect (6) on the 
stability. Elements of the iron group accelerate decomposition ; 
the presence of magnesium and aluminium increases the 
instability when the alkalinity is low. Elmanowitsch and 
Zaleski(4) found that acids increased, while alkalies decreased, 
the amount of (calcium) hypochlorite decomposed (at boiling 
temperature) by natural waters. Such factors as temperature, 
alkalinity, and the presence of phenols and hydrogen sulphide 
promote the decomposition of hypochlorites in sewage. (5) 

The accelerating influence of a high organic content in water 
on the decomposition of hypochlorites has been much studied. 
Asparagin, peptone and allied products, (8) albumin and its decom- 
position products, (4) sewage, (5) urine, sweat, saliva, and body 
products in general (10) have an especially great effect on the 
chlorine consumption. Apparently there is no direct paral- 
lelism between oxygen-consuming capacity (as measured by the 
reduction of permanganate solution) and chlorine-consuming 
power, (14) this no doubt due partly to the well-known inaccur- 
acy of determinations of oxygen-consuming capacity in regard 
to both quantity and kind of organic matter present in water, 
partly to specific interaction between the substances in water 
with chlorine. 

At first the disappearance of available chlorine from chlorin- 
ated water or sewage is very rapid, but it soon becomes slow. 
The reaction proceeds as though there were present, in some wa- 
ters at least, substances which are so readily oxidized by chlorine 
that they take it up before it can destroy the bacteria present. 
Although the bactericidal action of chlorine occurs simultan- 
eously with the chemical decomposition, the latter might pro- 
ceed rapidly enough greatly to impede or even to nullify the 
former. After the chlorine is once destroyed, bacterial growth 
might proceed unchecked, which furnishes a possible explana- 
tion for the repeatedly noted phenomenon — the great increase 



xii. a, i Heise: Chloride of Lime 19 

in bacterial content of waters in municipal distribution systems 
at varying periods after chlorination.(i5) 

Much work remains to be done on the substances formed 
by the interaction of hypochlorites with the organic matter 
found in water. It is because of these substitution pro- 
ducts (16, 20) that chlorinated waters used for public supplies 
often retain a peculiar chlorine odor and taste after all chem- 
ical trace of free chlorine has disappeared. In such cases — (21) 

these tastes and odours are not occasioned so much by chlorine or hypo- 
chlorites themselves as by inorganic and organic chloramines, and possibly 
other chlorine-substituted compounds formed by interaction with the 
organic matters present in waters. * * * They [the chloramines] are 
all germicidal, and all possess a more or less disagreeable odour. * * * 
He [Rideal] proved that chlorine * * * was working by substitution 
for hydrogen in ammonia and organic compounds, yielding products [chlora- 
mine and hydrazine] of higher germicidal power than chlorine itself. 

Race (19) found that when ammonia was added to hypo- 
chlorites, in the proportion of 1 part ammonia to 2 parts of avail- 
able chlorine, the germicidal action was increased threefold. 

While the addition of alum causes an immediate reduction 
in available chlorine, it has no apparent effect on the bacte- 
ricidal properties of the solution for twelve hours, according 
to the findings of Avery and Lye. (l) 

With waters high in organic matter it is probable that the 
main chemical change follows the course of a monomolecular re- 
action as in the case in the decomposition of hypochlorites in 
distilled water. However, there are, apparently, so many simul- 
taneous reactions that the accurate determination of the veloc- 
ity constant is not easy. From the fact that, in general, the 
amount of chlorine consumed by sewage in a given time is pro- 
portional to the amount of chlorine added, Glaser(5) concluded 
that the reaction is monomolecular. Race (18) followed the course 
of the reaction between hypochlorites and a colored water and 
came to the same conclusion. 

The effect of light, though a most important factor in the 
decomposition rate of hypochlorites, has been frequently over- 
looked or at least insufficiently emphasized, so that some of the 
work which has been done is not conclusive. The importance of 
this factor in the sterilization of swimming pools has already 
been discussed. (10) 

EXPERIMENTAL PART 

Determination of available chlorine. — For the work recorded 
in this paper, chlorine was quantitatively determined in the 



20 The Philippine Journal of Science im 

usual manner by titration with standard sodium thiosulphate 
solution in the presence of phosphoric acid, potassium iodide, 
and starch solutions, care being taken to keep such factors as 
temperature and concentration as uniform as possible to en- 
sure comparable results. Phosphoric acid was used instead of 
hydrochloric or sulphuric acid, because errors in titration due 
to the presence of ferric iron and chlorates are thus avoided, (li) 
The following graduation (4) of this determination, on the basis 
of the color of the iodine-starch reaction in the presence of 
hydrochloric acid, serves very well for the estimation of very 
small quantities of chlorine in 200 cubic centimeter samples: 

Doubtful, when CI content, is 0.11 mg. per liter. 
Barely noticeable, when CI content is 0.12 mg. per liter. 
Noticeable, when CI content is 0.13 mg. per liter. 
Weak, when CI content is 0.17 mg. per liter. 
Distinct, when CI content is 0.24 mg. per liter. 
Sharp, when CI content is 0.4 mg. per liter. 

The decomposition rate of calcium hypoclorite solution in the 
dark. — In order to study the decomposition rate of hypochlorites, 
known quantities of standard filtered chloride of lime solution 
were added to water in large glass jars provided with ground 
glass covers. Aliquot portions were pipetted off from time to 
time and analyzed. 

Whether the reaction was followed in distilled water, natural- 
Cartesian or river) waters, in solutions of organic substances, or 
in sewage, the decomposition curve had the same general trend. 
At first there was a rapid disappearance of chlorine, followed 
by an abrupt slowing down of the reaction. This sudden change 
of velocity generally occurred within from thirty minutes to one 
hour, after which the reaction proceeded very slowly, but very 
regularly, showing no sign of reaching an end point, though 
followed for weeks. Therefore it appears probable that the 
main reaction between chloride of lime and dissolved substances 
quickly approaches completion, after which the curve of gradual 
decomposition in water is followed. The experimental data se- 
cured are shown in Table I. The results for sewage were the 
most striking, and as they are apparently typical, on a large 
scale, of the reactions occurring in the other liquids, they are 
shown graphically in fig. 1. 

For the sake of completeness, the data for distilled water are 
plotted in fig. 2 for the entire eight hundred ninety hours during 
which the reaction was under observation. The curve appears 
to be typical for all of the reactions in the table. Though the 



Heise: Chloride of Lime 



21 



reaction was carried on in a small dark room the daily temper- 
ature variation of which was probably less than 3° C, the 
temperature change no doubt affected the results somewhat. 
Experimental errors appear somewhat exaggerated on the curve, 
because of the large scale on which chlorine concentrations are 









Time in min 

a 20 30 *0 SO SO 70 8 


ltes. 
9 


ICO HO 120 130 14-0 


210 
200 
190 
ISO 




| 
























1 






















































J 
























































































c iSo 

o 

5 
































































a 
































1 
no 

c 






























































1 
































1 
































80 

'0 




























































































SO 






























































































1 
















Zj 


— — 











Fig. 1. The decomposition of chloride of lime in sewage in the dark. 

shown. The abrupt change in velocity indicated in fig. 1 does 
not appear in fig. 2, owing to the small scale on which time is 
plotted in the latter. 



22 



The Philippine Journal of Science 



Table I. — Decomposition of chloride of lime in the dark.' 







s5 




















Time. 


■S,; £ 


















< 


Hrs. mins. 










b37.2 





6 


36.0 





11 


35.5 





15 


35.1 


1 


14 


34.8 


5 


43 


34.4 


23 


34 


33.1 


29 


15 


33.0 


48 





32.5 


70 





31.0 


98 





30.6 


194 





29.6 


890 





26.6 



IN 5 LITERS DIS- 
TILLED WATER 
PLUS 0. 1 GRAM 
AMMONIUM 
ACETATE. 



IN SEWAGE. 



Time. 


i" 

O OS o 

hi 

>X p. 
< 


Hrs. mins. 

6 
12 

37 
3 30 
5 
5 35 
22 
46 
73 
145 


t>36.0 
35.0 
34.2 
30.8 
30.7 
5.00 
4.75 
4.47 
3.78 
3.15 
2.8 









b38. 4 





4 


37.9 


o 


10 


34.2 





16 


34.2 





54 


33.9 


5 





33.5 


23 





32.7 


30 





32.4 


54 





31.5 


70 





31.0 






t>37. 


4 


9.6 


9 


4.5 | 


17 


2.7 


55 


2.4 


3 46 


1.9 


3 51 


2.0 


23 


1.4 



! 

Time. 


1 Available chlo- 
| rine. Parts 
1 per million. 






Hrs. mins. 










b228.0 





2 


116.0 





6 


100. 0(?) 





9 


38.0 





15 


35.2 





49 


30.5 


21 


20 


22.0 


45 





14.5 








213.0 


o 


2 


190.0 





4 


176.0 





5 


161.0 





7 


147.0 





8 


127.5 


o 


11 


102.0 





13 


78.6 





15 


65.5 





17 


55.8 ! 





20 


41.2 





22 


37.2 





25 


35.5 





28 


35.4 





32 


33.3 ; 


1 


1 


29.4 


1 


11 


28.7 


2 


29 


20.7 



alts were verified by R. H. 



molar oxalic acid solution 



■ In this and subsequent tables some of the experimental r 
Aguilar, chemist, Bureau of Science. 

b Calculated. 

c The decomposition of hypochlorite proceeded so rapidly in 
that it could not be followed quantitatively. 

Table II. — Decomposition of a solution of chloride of lime in tap water 
in diffused daylight. 



Time. 


Available 
chlorine. 

Parts per 
million. 


Time. 


Available 
chlorine. 
Parts per 
million. 


Hrs. mins. 



2 

30 

2 00 


0.8 
0.6 
0.3 
trace 


Hrs. mins. 

2 

45 
1 25 
3 45 


10.0 
9.7 
8.5 
7.9 
7.0 







Heise: Chloride of Lime 

Time in hours. 



23 



1 .iter TCit i 

«j „ ***~' -»». 



FIG. 2. The decomposition of chloride of lime in distilled water in the dark. 

The data at hand for the progressive decomposition of chloride 
of lime do not show clearly the order of reaction. The decom- 
position in sewage is probably a heterogeneous reaction. (17) 
Fig. 1 shows that for a time the reaction velocity was practically 

constant (^=# ; 1j:=|). 

The decomposition rate of calcium hypochlorite solution in the 
light. — In the light the decomposition more nearly approached 
the typical rate of a monomolecular reaction. The data obtained 
in a typical series of determinations are given in Table II and 
are plotted in fig. 3. Comparison with fig. 1 shows clearly the 
difference between the course of the dark and the light reaction. 
As indicated in the curve, the photochemical decomposition pro- 
ceeded in a regular manner. As the experiments were performed 
in the diffused light of the laboratory, the light was a variable 
factor and caused deviations from the true course of the pho- 
tochemical decomposition. 

It is interesting to note in this connection that in measurements 
of the decomposition of chloride of lime in different waters in 
the dark and in the light the differences between the amounts 
decomposed in the light and in the dark were very uniform for 
any given light intensity, regardless, within the limits observed, 
of the specific chlorine-binding power of the various waters used. 
This is evident from the data in Table III. The last column 
shows that for any one day the differences between the "light" 
and "dark" determinations were nearly the same, though the 
waters under observation varied widely in their ability to de- 
compose chloride of lime in the dark. 



24 



The Philippine Journal of Science 



Time in minutes. 

100 120 1*0 



ISO 200 120 



< 



y- • — < 1 1 

k 1 — ' • — i — I — i 

\ 

-^ J i I ; ti- 



Fig. 3. The decomposition of chloride of lime solution in diffused daylight. 

Table III. — Effect of light on chlorine consumption; 200 cubic centimeter 
samples of different waters digested two hours at 28° C. 



Available chlo- 
rine consumed. 




[n dif- 
fused 
day- 
light. 


In 
dark. 


Differ- 
ence. 


mg. 


mg. 


mg. 


0.28 


0.11 


0.17 


0.47 


0.25 


0.22 


0.14 


0.14 


0.00 


0.45 


0.45 


0.00 


0.50 


0.24 


0.26 


0.66 


0.45 


0.21 


0.65 


0.23 


0.42 


0.87 


0.43 


0.44 


0.53 


0.23 


0.30 


0.60 


0.28 


0.32 



Sky was heavily overcast, and laboratory was dark. 



Heise: Chloride of Lime 



25 



The interaction of chloride of lime and organic matter as 
affected by concentration. — If the interaction of chloride of lime 
with organic matter is a monomolecular reaction, it follows di- 
rectly that the percentage of chloride of lime decomposed should 
be independent of the initial concentration. In other words, the 
amount of chlorine consumed in a given time should be a con- 
stant fraction of the quantity added. For example, (5) when 
the addition of available chlorine to sewage is doubled, the amount 
of chlorine absorbed will also be doubled, within certain limits, 
a fact which has an important bearing on the economical appli- 
cation of hypochlorites for disinfecting purposes. 

A good illustration of this regularity is shown in the following 
series of determinations on the interaction of chloride of lime 
and urea. Different quantities of filtered chloride of lime solu- 
tion, varying in strength from 2 to about 100 milligrams of 
available chlorine, were diluted to 100 cubic centimeters, placed 
in glass-stoppered bottles, and digested twenty-one hours in the 
dark at 28° C. with 2 cubic centimeters of 0.01 molar urea solu- 
tion. At the end of the digestion period the samples were anal- 
yzed, with the results listed in Table IV. 

Table IV. — Interaction of varying concentrations of calcium hypochlorite 
with urea. 



Hypochlorite add- 
ed as— 


Available chlorine 
consumed. 


Solution. 


Available 
chlorine. 


Ob- 
served. 


Calcu- 
lated, a 

1 


cc. 


mg. 


mg. 


1 
mg. 


1.0 


2.3 


2.1 


2.2 


1.5 


3.4 


3.1 


3.3 


2.0 


4.6 


4.3 


4.4 


3.0 


6.9 


6.5 


6.6 


4.0 


9.2 


8.8 


8.9 


5.0 


11.5 


11.0 


11.1 


6.0 


13.8 


13.3 


13.3 


7.0 


15.3 


14.9 


15.5 


8.0 


18.3 


17.7 


17.7 


9.0 


20.6 


20.0 


20.0 


10 


22.9 


22.2 


22.2 


15 


34.4 


33.5 33.3 


20 


45.8 


44.5 44.4 


25 57. 2 


55.6 55.5 


30 68. 7 


67.0 ' 66.6 


40 91. 6 


88.7 | 88.8 i 



From the values obtained with 6, 8, 9, and 10 cubic centimeters of solution. 



26 



The Philippine Journal of Science 



Almost without exception the experimental results vary from 
the calculated values by quantities well within the range of ex- 
perimental error. Apparently this regularity holds good for 
wide ranges of concentration with many different substances. 2 
Even with fairly pure natural waters similar results may be 
obtained under certain conditions. Thus 100 cubic centimeter 
samples of a clear river water, boiled with varying amounts of 
filtered chloride of lime solution for fifteen minutes in diffused 
daylight, reacted as follows: 

Table V. — The interaction of a river water and varying quantities of 
chloride of lime solution at 100° C. 



Chloride 

Chloride of lime of lime 

added as — Iconsumed 



mg. 

I 0.52 

J 0.49 

| 0.96 

| 0.84 

I 1.68 

1 1.82 



The ratio of the amounts of chlorine consumed is approximately 
the same as that of the amounts added, namely, 1:2:3. 

With sewage there is a similarity of reaction, with some curious 
differences. The percentage of chlorine consumed may increase 
with increasing additions of chloride of lime, giving rise to the 
peculiar phenomenon that the available chlorine left is greater 
after digestion with small quantities of chloride of lime than 
after digestion with much larger quantities for an equal period 
of time. This observation was repeated too frequently to be cap- 
able of explanation on the ground of experimental error, though 
some series showed these abnormalities to a lesser degree than 
others. 

This irregularity in chlorine consumption is also evident in 
the data of previously published work. For example, Glaser (5) 
says in effect: 

As a matter of fact, there are at hand several experiments which deviate 
from the regular order, for instance, the progressive decomposition with 

"Avery and Lye (1) found that upon the addition of alum to water the 
available chlorine was apparently reduced, within certain limits, in direct 
proportion to the amount of alum added. 



XII. A, 



Heise- Chloride of Lime 



27 



a concentration of chloride of lime of 1:1000 recorded in tables 28 and 29 
' * [but] * * * a similar marked decrease in available chlorine 
could not be detected on repeating the experiments with the same concen- 
tration. It should be pointed out that there are other small deviations from 
the theoretical quantities of chlorine. 

The analytical data given in Table VI show clearly that at 
certain concentrations the reaction is completely changed and 
that the irregularities noted by Glaser were not due to experi- 
mental error. 

Table VI. — Decomposition of chloride of lime in fresh sewage * {concen- 
tration of chloride of lime expressed as milligrams of available chlorine). 

A. EFFECT WITH VARYING PERIODS OF DIGESTION. 



| 18.75 



7.40 
7.17 

7.24 



1.14 
1.24 
1.37 



11.35 
11.58 
11.51 



8.06 
12.95 
13.20 
13.25 



ft - -™ e " d - Ad H Left ~ s "" e " d 



2.93 

2.76 
2.82 
0.48 
0.78 
0.38 
4. 11 
3.55 
4.03 



1.87 
8.90 
8.60 
9.00 
14.64 
15.20 
14.72 



1.17 
2.34 
4.69 
7.03 
9.38 
11.71 
14.09 
16.43 
18.75 
21.1 
23.4 
25.8 
28.1 
37.5 



15. 



10.5 
13.6 
15.6 

17.4 



20.2 
20.6 

21.7 



3 Sewage, 50 cubic centimeters ; temperature of digestion, 28° C. 
" Time of digestion, forty-five minutes. 
* Time of digestion, one hundred minutes. 
d Time of digestion, eighteen hours. 

B. EFFECT WITH SEWAGE SAMPLES OF DIFFERENT AGES. 



Series 5. Fresh sewage. 



Series 6. Same sample, after 24 hours. 



2.20 
4.40 



35.2 
70.4 



0.07 
0.47 
0.95 



1.03 

1.73 
3.45 
8.56 
15.9 
19.4 



5.00 
6.00 i 



40.0 
50.0 





0.26 

0.42 

0.64 

1.10 

1.90 
1.35 
1.00 
15.8 
25.2 
33.5 



3.90 
4.80 



Per cent. 
100 
87 
86 
84 
78 



28 The Philippine Journal of Science I9t7 

Table VI. — Decomposition of chloride of lime in fresh sewage (concen- 
tration of chloride of lime expressed as milligrams of available 
chlorine ) — Continued. 

C. EFFECT WITH CONSTANT VOLUME OF LIQUID (CONCENTRATIONS OF CHLO- 
RIDE OF LIME EXPRESSED AS PARTS PER MILLION OF AVAILABLE CHLORINE). 



Series 8. 



; Added. | Left. j Consumed. | Left. ; 



4.6 1.9' 2.7 69 



9.2 



3.7 | 5.5 ; 60 \ 3.2 j 

18.4 I 7.3 j 11.1 J 60 j 10.6 

27.5 13.7 : 13.8 50 '' 13.1 
45.8 j 16.4 ' 29.4 j 64 ! 28.0 
64.2 I 5.2 ' 59.0 ! 92 I 27.5 ! 
82.5 3.5 ; 79.0 ; 96 i 8.5 | 



184 j 41.0 j 143 I 78 i 21.1 

275 j 108 I 167 [ 61 j 91.1 

368 i 197 j 171 j 47 j 171 

460 ! 249 ! 211 i 46 | 258 

690 ' 445 ; 245 , 36 j 472 

920 707 213 ] 23 ' 690 

1,835 I 1,385 ; 450 j 25 j 1,415 

2,750 j 2,093 | 657 j 24 j 1,925 

4.580 3,615 | 965 ; 21 3,330 





Per cent. 


4.0 


87 


6.0 


65 


7.8 


42 


9.4 


34 


17.8 


39 


36.7 


57 


74.0 


90 


106 


96 


132 


95 


163 


88 


184 


67 


197 


54 


202 


44 


218 


32 


230 


25 


420 


23 


825 (?) 


30 


1,250(?) 


27 







The foregoing data show the peculiar fact that with increasing 
additions of available chlorine the amount left after digestion 
first increased, next decreased, and finally again increased. For 
certain concentrations the available chlorine left after digestion 
was greater with small additions of chlorine than with much 
larger ones. The different series listed above were obtained on 
different days with different samples of sewage, so that the 
agreement between the results recorded is as good as is to be 
expected. As a matter of fact, different samples vary greatly 
and the same samples change with time, as indicated by the 
differences between series 6 and series 7. 

Owing to the change in the volume of the digested liquid, 
caused by large additions of chloride of lime solutions, discrep- 
ancies in the analytical results may occur. In series 8 these 
deviations were avoided by keeping the quantity of liquid constant 
by the addition of enough water to the first members of the series 
to compensate for the differences in volume caused by the ad- 
dition of chloride of lime solution. 

Typical series in the foregoing table are graphically shown in 



Heise: Chloride of Lime 



29 



fig. 4, the experimental data from Glaser's(5) Table XXIX for 
20-hour digestion periods being inserted for comparison. For 
use in this figure, the data in Table VI were recalculated as parts 
per million. 



Available chlorine added. Parts per million. 



.*UC 2tt>l 





i 1 ; ' 


















o 








' 
























c 






























^ 




'* 








































































1 






/, 


















JrV 






^ 






















£tA— 




***s 


























^_Jf^-^'°~^ 


-%- 


^ 
























fl 




























if ] 



























Fin. 4. The decomposition of chloride of lime in sewage in the dark. Curve a, results 
of Glaser ; curve b, Table VI, series 4 ; curve c, Table VI, series 7 ; curve d, Table 
VI, serie, 8. 

Effect of the concentration of substances reacting with chloride 
of lime. — The more concentrated the solution with which the 
chloride of lime reacts, the more chlorine is used up. This is 
to be expected, and it has been pointed out so often that it needs 
no further proof here. There is, however, a curious anomaly 
which may occur and which may lead to confusion in the inter- 
pretation of the ordinary laboratory tests of the chlorine-binding 
power of a water. The interaction of urea and chloride of lime 
in the dark at moderate temperature furnishes an illustration 
of this, when the decomposition of chloride of lime is measured 
in the usual manner, that is, by means of standard sodium 
thiosulphate, in the presence of potassium iodide, phosphoric acid, 
and starch. 

In Table VII are the results obtained when the same quantity 
of chloride of lime was allowed to react with different quantities 
of urea. They show that, for very low concentrations of urea, 
increase in concentration leads to increased chlorine consump- 
tion ; with higher concentrations, however, the chlorine con- 



30 



The Philippine Journal of Science 



sumption decreases. The results are also shown graphically in 
fig. 5. 

The scope of the work does not warrant a detailed study of 
this reaction :! at this time. The results are presented because 
they are sufficiently definite to show the disturbing factors that 
may be present in the tests of the chlorine consumption of a 
water or sewage. Probably different chlorine-substitution pro- 
ducts are formed with different concentrations of urea, some 
of which react immediately with potassium iodide, liberating 
iodine, some reacting slowly or not at all, so that the amount of 

Urea additions in milligrams. 





o 


3. 




. 




































1 














































































































































































































































































































































































































































































10. E 


. T 


he c 


oci m 


IPOS 


tion 


of 


:hloi 
t 


ide 

ons 


.f ii 


me 

the 


n u 

JarL 


ea 


olut 


ons 


of N 


aryi 


ng concentra 



available chlorine left in solution may not be a true measure of 
the reaction. Since these products may have even stronger 
disinfecting action (21) than chloride of lime, the determination 
of chlorine consumed is not necessarily an index of the quantity 
of hypochlorite required for disinfection. 

According to Hairi,(8) the liquids in which the disinfecting 
action of chlorine is greatly retarded show a strong chlorine- 
consuming power. In a solution of peptone, for example, neither 
the chlorine consumption nor the inhibition of germicidal action 
varies in the same manner as the concentration. 



'After this work had been completed, I found that Dakin (3) had 
reported a similar anomalous reaction between hypochlorous acid and 
sheep serum. 



Heise: Chloride of Lime 



31 



Table VII. — Interaction of chloride of lime solution with varying concen- 
trations of urea. 

[Water, 100 cubic centimeters: urea, additions as noted. Five cubic centimeters of chloride 
of lime solution digested eighteen hours in the dark at 28°C] 

SERIES A. ADD1T1ON-11.0 MILLIGRAMS AVAILABLE CHLORINE. 



Urea. 


Available chlorine. 


Solution 
added. 


Weisht 
added. 


Consumed. 


Left. 


cc. 


0.1 
0.25 
0.50 
1.00 


g. 



0.003 

0.007 

0.015 

0.030 


mg. 

0.05 
10.3 
10.8 (6) 
10.8 (4) 
10.8 


mg. 
10.9 (5) 
0.7 

0.1 (4) 
0.1 (6) 
0.2 

.1 



SERIES B. ADDITION— 11.8 MILLIGRAMS AVAILABLE CHLORINE. 









0.1 


11.7 


0.5 


0.01 (5) 


11.5 (5) 


0. 1 (5) 


1.0 


0.03 


11.5 


0.3 


2.0 


0.06 


11.4 


0.4 


3.0 


0.09 


11.3 (5) 


0.4 (5) 


4.0 


0.12 


11.3 


0.5 


6.0 


0.18 


11.2 


0.6 


7.0 


0.21 


11.1 (5) 


0.6 (5) 


8.0 


0.24 


11.1 


0.7 


9.0 


0.27 


11.0 


0.8 


10.0 


0.30 


10.8 


1.0 


11.0 


0.33 


10.6 (5) 


1.1 (5) 


12.0 


0.36 


10.4 


1.4 


13.0 


0.39 


10.3 


1.5 


14.0 


0.42 


10.1 


1.7 


15.0 


0.45 


9.9 


1.9 


16.0 


0.48 


9.8 


2.0 



GENERAL DISCUSSION 



Since there is no direct parallelism between the oxygen-con- 
suming capacity and the amount of chlorine a water or sewage 
is capable of taking up, the determination of chlorine-consuming 
capacity, in spite of its limitations, is an important test in the 
laboratory control of hypochlorite disinfection. Hairi(8) di- 
gested 100 cubic centimeter samples with chloride of lime solu- 
tion in excess (13 milligrams available chlorine) and titrated 
after an hour in weakly acid solution. If digestion is carried 
out at constant temperature in the dark, this procedure will give 
concordant, relative values for the chlorine absorption. 

Elmanowitsch and Zaleski(4) recommended the determination 



32 The Philippine Journal of Science im 

of chlorine consumption at boiling temperature, with carefully- 
regulated constant heating, for fifteen-minute digestion periods, 
with lime water added to the water under examination. The 
authors obtained very concordant results by this process. The 
figures given in Table V were obtained without addition of lime 
water, but show that the method is accurate enough for labo- 
ratory use. However, this method is open to objection, because 
it does not take into account conversion of hypochlorite to chlo- 
rate at boiling temperature. 

Digestion in the dark at room temperature is so easily carried 
out, and gives such uniform results if care is used in keeping 
conditions constant, that it is to be preferred to the methods 
employing heat. 

Whether chloride of lime shows fluctuations in bactericidal 
effect similar to those noted for its chemical decomposition 
(Table VI, fig. 4) is not clear. The results of Glaser(5) indicate 
that it does not, but these observations were made with such high 
concentrations of chloride of lime that they are not conclusive. 

It is true that the data under discussion were obtained in ex- 
periments on sewage, but apparently similar fluctuations may 
occur with ordinary water as well and with small hypochlorite 
concentration. Thus Stokes and Hachtel(22) in their work on 
the treatment of Baltimore spring water by calcium hypochlorite 
report that — 

when 1.5 parts of available chlorine per million parts of water were used 
there is practically no residual chlorine in the water. In one case when 
1.75 parts were used there was a large amount of residual chlorine, giving 
an average .... of .574. On another occasion, however, when this amount 
was used the residual chlorine was less, giving an average of 0.24 per 
million parts of water. When 2.0 parts of available chlorine were used 
there was an average ... of 0.206, and when 2.5 parts of available chlorine 
were used for treatment there was an average of 0.62 parts per million 
parts of water. These results, therefore, are somewhat variable, and it 
is hard to explain the greater amount of residual chlorine when 1.75 parts 
were used than when 2.0 parts were used. 

From the evidence at hand it is clear that the ordinary pro- 
cedures for determining the amount of interaction between hy- 
pochlorites and waters and sewage are influenced by so many 
factors that, unless very carefully interpreted, they are very 
likely to be misleading. If it were only the available chlorine 
that had germicidal action, the analytical control of hypochlorite 
sterilization would be relatively simple. Even in this case, how- 
ever, the quality of a water would be an important consideration, 
since waters differ widely in their ability to liberate chlorine 
from hypochlorites. (21) Thus two different waters brought to 



xii. a. i Heise : Chloride of Lime 33 

this laboratory were treated with chloride of lime solution and 
immediately titrated with sodium thiosulphate in the usual man- 
ner. Acidified, 100 cubic centimeter samples showed 6.5 milli- 
grams of available chlorine, whereas unacidified samples showed 
only 2.4 and 4.0 milligrams, respectively. Since the amount of 
chlorine taken up by water or sewage depends upon the amount 
and concentration of chlorine added, on the temperature and 
time of digestion, on light, and on the quality of the liquid 
studied, it is obvious that, at best, only relative data, available 
for laboratory control of disinfection problems, are obtained by 
the methods ordinarily employed. Such determinations do not 
necessarily give any indication of the presence of germicidal 
products formed by the interaction of hypochlorites and the in- 
gredients of waters and sewage ; hence they lose much of their 
significance for disinfection problems unless they are studied in 
conjunction with bacteriological data. In all cases, the chlorine 
consumption should be determined as nearly as possible under 
the same conditions of temperature, illumination, and concen- 
tration that obtain in actual practice. 

SUMMARY 

The decomposition rate of chloride of lime in water, sewage, 
and solutions of organic substances was studied. In the dark, 
at 28° C, the reactions proceeded with almost constant velocity 
for periods of thirty minutes to one hour, after which they pro- 
ceeded very slowly. In the light the decomposition rate was 
greatly accelerated. 

In general, the amount of available chlorine consumed is pro- 
portional to the concentration in which it is added, as shown by 
the interaction of chloride of lime and urea solution. However, 
for certain definite concentrations of sewage this regularity 
fails. 

A study of the reaction between chloride of lime with varying 
quantities of urea showed that the chlorine consumption, as meas- 
ured by the starch-potassium-iodide reaction, is not necessarily 
proportional to the concentration of organic matter. 

The determination of the chlorine consumption of a water or 
sewage, though of importance in the control of hypochlorite 
disinfection, is not sufficient in itself and should be supplemented 
by bacteriological tests. 



34 The Philippine Journal of Science 1m 

REFERENCES 

(1) Avery, C. R., and Lye, 0. G. Ann. Rep. Prov. Bd. Health Ontario 

(1914), 33, 150-155; Exp. Sta. Rec, 34, 885, through Chem. Ab. 
(1916), 10, 2780. 

(2) Bhaduri, I. Zeitschr. f. anorg. Chem. (1897), 13, 385. 

(3) Dakin, H. D. Brit. Med. Journ. (1916), 852-854. 

(4) Elmanowitsch, N., and Zaleski, J. Zeitschr. f. Hyg. (1914), 78, 

461. 

(5) Glaser, E. Arch. f. Hyg. (1912), 77, 165. 

(6) Griffin, M. L., and Hedallen, J. Journ. Soc. Chem. Ind. (1915), 

34, 530. 

(7) Grimm, N. Mitteilung aus der Priifungsanstalt fur Wasserversorgung 

u. Abwasserbeseitigung (1912), No. 16, 297, through Chem. Zentr. 
(1912), 83, II, 625. 

(8) Hairi, E. Zeitschr. f. Hyg. (1913), 75, 40. 

(9) Heise, G. W. Phil. Journ. Sci., Sec. A (1916), 11, 1-13. 

(10) Heise, G. W., and Aguilar, R. H. Phil. Journ. Sci., Sec. A (1916), 

11, 114. 

(11) Kedesdy, E. Mitt. kgl. Materialpriifungsamt (1914), 32, 534, through 

Chem. Ab. (1915), 9, 2044. 

(12) Koch, R. Mitt. a. d. kaiserl. Gesundheitsamte (1881), 1, 234-282. 

(13) Lewis, W. C. McC. Journ Chem. Soc. (1912), 101, 2371. 

(14) Lode. A. Arch. f. Hyg. (1895), 24, 236-264. 

(15) Longley, F. F. Am. Journ. Pub. Health (1915) 5, 918. 

(16) Meisenheimer, J. Ber. d. deutsch. Chem. Ges. (1913), 46, 1148. 

(17) Mellor, J. W. Chemical Statistics and Dynamics. Longmans Green 

& Co., London (1914), 124 et seq. 

(18) Race, J. Journ. Am. Water Works Assoc. (1916), 3, 439. 

(19) Idem. Can. Engr. (1916), 30, 345, through Chem. Ab. (1916), 10, 

1565. 

(20) Raschig, F. Chem. Zeitg. (1907), 31, 926. 

(21) Rideal, S., and Rideal, E. K. Water Supplies. D. Appleton & Co., 

New York (1915), 190 et seq. 

(22) Stokes, W. R., and Hachtel, F. W. Am. Journ. Pub. Health (1916), 

6, 1224-1235. 

(23) Traube, M. Zeitschr. f. Hyg. (1894), 16, 149-50. 



ILLUSTRATIONS 

TEXT FIGURES 

Fig. 1. The decomposition of chloride of lime in sewage in the dark. 

2. The decomposition of chloride of lime in distilled water in the dark. 

3. The decomposition of chloride of lime solution in diffused daylight. 

4. The decomposition of chloride of lime in sewage in the dark. Curve 

a, results of Glaser; curve b, Table VI, series 4; curve c, Table 
VI, series 7; curve d, Table VI, series 8. 

5. The decomposition of chloride of lime in urea solutions of varying 

concentration in the dark. 

35 



A CHEMICAL INVESTIGATION OF THE SEEDS OF PANGIUM 
EDULE AND OF HYDNOCARPUS ALCALDE l 

By Harvey C. Brill 

(From the Laboratory of Organic Chemistry, Bureau of Science, 

Manila) 

In an article 2 describing the investigation of the seeds of 
Hydnocarpus venenata and the properties of the oil from these 
seeds, announcement was made of the continuation of this in- 
vestigation on the oils from seeds of various related plants 
growing in the Philippine Islands. The examination of such 
seeds was undertaken because of the fact that they are closely- 
allied to the genus Hydnocarpus, which is closely related to the 
genus Taraktogenos, in order to determine if they have properties 
similar to the seeds of the latter. Chaulmoogra oil is obtained 
from Taraktogenos, but without doubt much of the oil sold as 
chaulmoogra is obtained from Hydnocarpus, since the chemical 
properties of certain of these oils are practically identical with 
the properties of chaulmoogra oil, which would make substitution 
easy and practically impossible to detect. Data concerning the 
seeds of Pangium edule Reinw. and Hydnocarpus alcalde C. DC. 
are herewith presented. 

PANGIUM EDULE REINWARDT 

Pangium edule is indigenous to the Malayan Archipelago and 
is found in various parts of the Philippine Islands. The seeds 
are flattened; their average size is about 5 centimeters long by 
3 centimeters wide. They are embedded in a crustaceous peri- 
carp, which is about 22 centimeters long by 15 centimeters in 
diameter. Mature and immature seeds were examined. 

The Pangium edule seeds used by me were furnished by the 
Philippine Bureau of Forestry. 

De Jong 3 announces the isolation of a cyanogenetic glucoside 
from the leaves of Pangium edule that is identical in all re- 
spects with that found in Gynocardia odorata by Power et. al. 4 

1 Received for publication November, 1916. 

'Brill, Harvey C, This Journal, Sec. A (1916), 11, 75. 

3 De Jong, Awk., Recueil des travaux chimiques des Pays-Bas et de la 
Belgique (1909), 28, 24; ibid. (1911), 30, 220. 

4 Power, F. B., and Lees, F. H., Journ. Chem. Soc. London (1905), 87, 
349; Power, F. B., and Barrowcliff, M., ibid. (1905), 87, 896. 

37 



38 The Philippine Journal of Science mi 

and named gynocardin by them. As the glucoside is accom- 
panied by an emulsinlike enzyme, gynocardase, in the plant, any 
delay in working up the seeds or leaves entails a loss in the 
glucoside content, due to its hydrolysis by the enzyme. In order 
that this loss might be reduced to a minimum, instructions were 
given to the Bureau of Forestry rangers to heat the seeds in 
boiling water for one hour, insuring the destruction of the gyno- 
cardase, and then to dry them in the sun before delivery to the 
Bureau of Science. In spite of this precaution, in several cases 
an odor of hydrocyanic acid was noticeable when the seeds 
were received, showing that hydrolysis of the glucoside had taken 
place. These particular samples were not dry. Owing to the 
lateness of the rainy season and the prevalence of rain, it was 
impossible for them to be dried properly before shipment and 
they molded somewhat in transit. Dox ' has shown that molds 
formulate all the known enzymes, regardless of the character 
of the substrase; consequently the small amount of glucoside 
present in some of the seeds received may have been due to the 
action of enzymes formed in this manner. As stated, both ma- 
ture and immature seeds were received and examined. The 
mature seeds in no case gave any large amount of hydrocyanic 
acid when tested for the presence of gynocardin. The failure 
of strong positive tests cannot be attributed to the hydrolysis 
by molds in every case, since of the several samples of mature 
seeds examined not all were molded. It would appear that the 
amount of this glucoside decreases as the seeds ripen. A de- 
crease in the quantity of glucoside would appear plausible if one 
holds to the theory that its function is to sterilize any injury 
received by the fruit and thus prevent further injury from the 
introduction into the wound of molds and bacteria. With the 
maturity of the fruit the need for this protection would cease 
to exist. 

Upon receipt, the seeds were immediately shelled, placed in the 
oven and dried, and then ground. In some cases the oil was 
removed by extraction in a large syphon-extraction apparatus, by 
means of petroleum ether ; in others most of the oil was removed 
by expression as is done in the preparation of amygdalin from 
almonds ; in some others the original ground seeds were extracted 
with alcohol, which extracted only a small amount of the oil, but 
removed the glucoside. By evaporation to dryness and extrac- 
tion of this residue with ether the oil could all be removed and 
the black gummy residue treated for the isolation of gynocardin. 

Dox, A. W., Plant World (1912) , 1 5, 40. 



xii, a, i Brill: Pangium edule and Hydnocarpits alcaUe 39 

Where immature nuts were being handled, the most successful 
method for their treatment was the following : The ground nuts 
were triturated with hot water, allowed to stand for some time, 
expressed, and the process repeated. The water was removed 
by distillation in a partial vacuum, the black gummy mass was 
triturated repeatedly with hot 90 per cent alcohol, and the alcohol 
was removed by distillation in a partial vacuum or more slowly 
at not too high a temperature at atmospheric pressure. The 
residue was repeatedly extracted, this time with hot absolute 
alcohol, and ether was added to the extract until a precipitate 
no longer formed. The extract was evaporated to dryness, and 
the residue was washed a number of times with acetone in the 
cold. In some instances this treatment left a semiviscous mass 
that could be crystallized from water. In a few instances the 
residue was dissolved in hot acetone and the acetone evaporated. 
However, gynocardin is not readily soluble in acetone, and as a 
small amount of fat persistently adheres to the glucoside and this 
is dissolved by the acetone along with the glucoside, dissolving in 
water preceded by washing with cold acetone was found to be 
more successful in obtaining a pure crystalline compound. Crys- 
tallization will not take place in the presence of a small amount 
of the oil; consequently its removal was necessary. The com- 
pound was obtained in the form of hairlike, golden yellow 
crystals, with a melting point of 160° C. The yield of the pure 
substance was between 0.2 and 0.3 per cent based on the weight 
of the dry kernels of the immature nuts. In another sample of 
immature seeds a quantitative estimation of the hydrocyanic 
acid was made by suspending 4 grams of the ground seeds in 
water and hydrolyzing them at a temperature of 39° C. with 
emulsin. At the end of forty-eight hours the hydrocyanic acid 
was distilled into sodium hydroxide containing a trace of potas- 
sium iodide and the latter was titrated to opalescence with 0.01 N 
solution of silver nitrate. This procedure indicated a content 
of 0.0126 per cent of hydrocyanic acid corresponding to 0.156 
per cent of gynocardin. 

PROPERTIES OF GYNOCARDIN 

The best known hydrocyanic glucoside is amygdalin. 8 Within 
recent years many new ones have been discovered, and their 
properties have been studied. The chemical properties of the 
cyanogenetic glucoside, gynocardin, have been noted by De Jong 

" Abderhalden, Emil, Biochemisches Handlexikon. Julius Springer, Ber- 
lin (1911), 2, 707. 



40 



The Philippine Journal of Science 



and by Power and Lees in the references cited. For purposes 
of identification, the melting point and specific rotation of the 
compound isolated by me were determined, and comparison was 
made with those of the products isolated by these works. 
Table I. — Melting point and specific rotation of gynocardin. 



SeedS „°JJ%™* ardia Leaves of Pangium edule. b I Seeds °JS™° ium 



Melting point. _°C-- 
Specific rotation in 
chloroform solu- 
tion. 



[ a ] 2 |j- (16. 885%) = +62. 2° 



* Powers and Lees, loc. cit. b De Jong, loc. cit. c Occurring in the Philippine Islands. 

The properties of the compound isolated from the seeds of 
Pangium edule are so nearly identical with those of the glucoside 
from the seeds of Gynocardia odorata and from the leaves of 
Pangium edule that no doubt exists as to their identity. De 
Jong found that the concentration of the chloroform solution 
had considerable influence on the specific rotation of the com- 
pound. This fact is apparent in the results quoted above from 
his publication. 

Gynocardin differs from the other members of this class of 
compounds in its marked stability in the presence of acid hy- 
drolyzing agents. It also hydrolyzed very slowly by emulsin. 



Table II.— 


■Hydrolysis of gynocardin by 


means 


of emulsin. 




' Quantity 




! 








1 of gyno- 
1 cardin. 
t- „ , Solution 
T,me. of2 

j grams in 
1 300 cc. 


Quantity 


Quantity 

of other 

sub- 


Glucoside 
hydro- 






emulsin. 


stance 
added. 


lyzed. 






1 water. 




i 








Min. ! cc. 


j 


Per cent. 






Grams. 






20 1 25 




none | 


0.68 






40 1 25 




none 


0.72 






60 


25 




none 


0.76 






20 


25 


0.10 


none 


0.96 






40 


25 


0.10 


none | 


1.02 






60 


25 


0.10 


none 1 


1.08 






20 


25 


0.05 


• 0.20 


0.52 






40 


25 


"0.40 


0.40 




















60 


25 




•0.30 


0.48 






20 


25 




t>l 


2.05 






40 


25 


0.05 


b5 j 


1.94 






60 


25 


0.05 
0.05 


no | 


0.52 




• Glucose 


in grams. 






b Hydroc 


fame ac 


id in cubi 


c centime 


ters of 0.1 


per cent 


solution. 



Brill: Pangium edule and Hydnocarpus alcaUe 



41 



Auld ' has determined the rate of hydrolysis for amygdalin 
by means of emulsin. Amygdalin reacts much more rapidly 
with emulsin as well as with acids than does gynocardin. His 
results with emulsin in the presence of added glucose and of 
added hydrocyanic acid are in accord with the results obtained 
by me for gynocardin, namely, the glucose inhibits the reaction, 
the amount being dependent on the relative quantity of glucose 
added, while hydrocyanic in smaller amounts accelerates the 
action, but when added in larger quantities the inhibiting effect 
is apparent. He has also found that the law of mass action 
does not hold good in its entirety for enzymic action. A large 
excess of the substrase does not cause a like increase in the rate 
of reaction when this addition has proceeded beyond a certain 
limit. The enzyme appears to be capable of reacting with only 
a certain amount of the substrase in a stated interval of time, 
and as long as the amount of this substrase exceeds a certain 
limit, no acceleration of the reaction takes place on the addition 
of more of the substrase. On the other hand, an increase in the 
amount of the enzyme gives a corresponding acceleration of the 
reaction, if the substrase is present in excess to react with the 
original amount of enzyme present. In the preceding table 
addition of more emulsin has accelerated the reaction. 

The juice from the crab contains a secretion which hydrolyzes 
amygdalin giving hydrocyanic acid. To examine gynocardin in 
this respect, a series of experiments was undertaken. The juice 
of crabs, purchased in the local market, was obtained by ex- 
pression, and comparative tests were made with it on amygdalin 
and gynocardin. 

Table III. — Action of crab juice on 1 per cent solutions of gynocardin and 
of amygdalin. 



Time. 


Glucoside. 


Gluco- 
side. 


Crab 
juice. 


Gluco- 
side 
hydro- 
lyzed. 


Hours. 
24 

24 
28 

: 

n 




cc. 
25 
25 
25 
25 
25 
25 


10 
10 
10 
10 
10 
10 


Per cent. 
51.99 
11.60 
69.10 
13.16 
75.67 
17.18 















According to Table III gynocardin is likewise attacked more 
slowly by crab juice than is amygdalin. Toluene was added in 



7 Auld., S. J. Manson, Journ. Chem. Soc. London (1908), 93, 1251. 



42 The Philippine Journal of Science 1917 

the above case for an antiseptic. The flasks containing the solu- 
tions were incubated at 39° C. 

In view of the hydrolyzing effect of crab juice it was thought 
that it might prove interesting to determine the effect of adding 
blood to similar samples. 

Table IV. — Action of blood on a 1 per cent solution of gynocardin. 



Time. 


Kind of blood. 


Blood. 


Gluco- 
Bide. 


Gluco- 

hydro- 
lyzed. 


Hours. 
21.5 
21.5 
22.5 
45.5 
45.5 
46.0 


Blood from apparently healthy person 


6 

3 
6 


cc. 
25 
25 
25 
25 
25 
25 


Per cent. 
1.45 
1.69 
3.63 
3.15 
6.14 
3.96 













« This blood was kindly furnished me by Dr. J. W. Smith, in charge of Bilibid Prison. 

The above results indicate that blood from tubercular pa- 
tients reacts more rapidly with gynocardin than does the blood 
of the normal person. The blood from the syphilitic person 
apparently reacted more rapidly the first day, but the sample 
incubated two days reacted no faster than the sample from the 
normal person. The Bureau of Science has collected some 
further data concerning this property when amygdalin is used 
that substantiates the above evidence. 

Amygdalin has been found to be harmless when taken into 
the system unless administered in the presence of emulsin. Gy- 
nocardin, when given in doses of 0.25 and 1 gram to guinea 
pigs, was found to be without any apparent effect. At the end 
of twenty-four hours no further observations were taken, as the 
pigs appeared perfectly normal. 

GYNOCAKDASE 

The enzyme gynocardase was obtained from the leaves of 
Pangium edule by grinding the leaves very fine and then pressing 
in a hydraulic press, adding an equal volume of alcohol to the 
juice and filtering. The precipitate was then ground under 
water with sand, filtered through cloth, and again precipitated 
by the addition of alcohol. The precipitate was nearly black 



xii. a, i Brill: Pangium edule and Hydnocarpus alcalx 



43 



and of gummy consistence. Determinations of its activity were 
made with amygdalin and gynocardin. 



Table V. — Activity of gynocardase on 1 per cent solutions of amygdalin 
and gynocardin. 



Time. 


Glucoside. 


En- 
zyme. 


Glucoside 
hydrolyzed. 


Hours. 
24 
24 
24 
24 


Amygdalin 

do 


cc. 

25 
25 
25 
25 


Grams. 
0.2 
0.2 
0.2 


Per cent. 
25.57 
26.02 
13.82 


Gynocardin . 









Here again gynocardin is less readily hydrolyzed than amyg- 
dalin. Since gynocardin is hydrolyzed by emulsin, which is a 
/3-enzyme, it must be a /3-glucoside. 8 

Gynocardase hydrolyzes both gynocardin and amygdalin ; con- 
sequently it must belong to the class of ,6-enzymes typified by 
emulsin. 

Several samples of nuts were examined for their oil content. 

Table VI. — Oil content and properties of oil obtained from the seeds of 
Pangium edule. 



Mature. Immature. 



Kernel content of air-dried nuts 

Dried kernel content based on air-dried nuts . 
Oil content based on dried kernels 



Per cent. Per cent. 
42. 67 36. 38 

29. 09 16. 28 

21.09 24.11 

I 



Melting point . 



Specific gravity 

Specific rotation in chloroform solution 

Iodine value (Hanus) 

Acid value cc. 0.1 N alkali 

Saponification value 

Index of refraction 



Free acids 

from oil of 

mature 

seeds. 



Clouds at 
18° C. 

0.9013 
+3.49 
113.5 



Shows some 
clouding at 
2°C. 

0.9049 

+4.28 

113.1 

0.52 

190.3 

1.4665 



Free 

acids 

from oil Oil from imraa- 
of imma 

ture 

seeds. 



No change 
at8°C. 



'Fischer, E., Ber. d. deutsch. chem. Ges. (1894), 27, 2985. 



44 The Philippine Journal of Science m? 

To determine the physiological activity of the oil, 2 grams 
and 2.5 grams, respectively, were administered per os to each 
of two guinea pigs. No ill effect was noted in either case. 

Owing to the limited quantities of the nuts available, no ex- 
haustive study of the oil could be made. However, enough was 
obtained to determine that palmitic and oleic acids and small 
quantities of an optically active acid are present. The latter 
may be either hydnocarpic or chaulmoogric or a mixture of the 
two. It resembles chaulmoogra and the hydnocarpus oils by the 
presence of an optically active oil. 

HYDNOCARPUS ALCALDE C. DE CANDOLLE 

A sample of the fruit of Hydnocarpus alcalas C. de Candolle 9 
was submitted by the Bureau of Agriculture to the Bureau of 
Science for identification and examination. It had been sent to 
the Bureau of Agriculture by Mr. T. Alcala, of Daraga, Albay 
Province, Luzon, under the local name dudu dudu. It was 
large, somewhat resembling a small unhusked coconut, about 
20 centimeters long and 15 centimeters in lateral dimensions. 
Within the pericarp were numerous seeds, measuring 4 centi- 
meters by 2.5 centimeters. 

Mr. Alcala writes: 

It is said that the oil extracted from the seeds is a good cure for wounds. 
It is generally believed to be poisonous, and when I ate six or eight of the 
boiled seeds I had a slight sickness; however, many children eat them raw 
without the slightest ill effect. 

At the suggestion of Mr. E. D. Merrill, botanist, Bureau of 
Science, that this fruit belongs to the Hydnocarpus family, an 
examination was made for hydrocyanic acid. The presence of 
this acid would indicate the existence of a cyanogenetic glucoside. 
All such tests resulted negatively. However, the fresh fruit 
or the unripe fruit might contain such bodies. Our examination 
of Pangium edule has shown that a decrease or complete disap- 
pearance of the glucoside results when the nuts age, unless the 
hydrolyzing enzyme is destroyed, and that with the ripening of 
the nut a decrease in the glucoside content probably occurs. Con- 
sequently the inability to obtain positive tests in this one sample 
of seeds for hydrocyanic acid is not to be considered as sufficient 
proof to warrant the statement that no cyanogenetic glucoside 
exists at any period in the growth of the fruit. 

'De Candolle, C, This Journal, Sec. C (1916), 11, 37. 



. i Brill: Pangium edule and Hydnocarpus alcalas 45 

Table VII. — Properties of nuts and oil of Hydnocarpus alcalas. 



Average weight of fruit 


grams 420 


Hulls 


per cent 59.68 


Seeds 


do 40.32 


Moisture in seeds 


do 60.50 


Oil in dry seeds 


do 65.50 


Melting point 


degrees 32 


Specific gravity at 30° C. 


0.9502 


Specific rotation in chloroform 


degrees +49.60 


Iodine value (Hanus) 


93.10 


Acid value cc. 0.1 N alkali 


3.90 


Saponification value 


188.90 


Index of refraction 


1.4770 


Reichert Meissl No. 


4.43 



The nuts and the extracted oil were administered per os to 
chickens without any noticeable ill effects. 

The free acids of the oil from Hydnocarpus alcalas were 
examined. 

Table VIII. — Properties of the free acids of the oil from Hydnocarpus 
alcalas. 

Melting point degrees 59 

Specific gravity at 30° C. 0.9342 

Specific rotation in chloroform 30/D degrees 53.65 

Iodine value (Hanus) 98.6 

Acid value cc. 0.1 N alkali 37.4 

Saponification value 193.0 

The free acids were then treated as described by Power for 
the separation of chaulmoogric acid. More than 90 per cent of 
these acids consist of a compound identical in properties with 
the acid designated as chaulmoogric by Power. 

Table IX.— Properties of the acid isolated from the oil from Hydnocarpus 

alcalas. 

Melting point degrees 68-69 

Specific rotation in chloroform 30/D do +59.69 

Iodine value (Hanus) 89.7 

0.3400 gram of silver salt gave 0.0950 gram of silver, equal to 27.94 per 

cent silver. 

0.5040 gram of sodium salt gave 0.1206 gram of sodium sulphate equal 
to 7.75 per cent sodium. 

Theoretical for C„H 3 COO Ag=27.86 per cent silver. 
Theoretical for C 17 H 3 COO Na=7.62 per cent sodium. 

Only a small amount of the acids remained in the mother 
liquor. No hydnocarpic acid could be isolated from this. The 
latter exists in very small quantities in Hydnocarpus alcalse if 
at all. Palmitic acid makes up the chief part of the remaining 



46 The Philippine Journal of Science 

portion with only traces of oleic acid. Both of these oils differ 
greatly in melting points from the oils from chaulmoogra and 
Hydnocarpus reported by Power and his coworkers. 

Pangium edule contains a large amount of olein and smaller 
quantities of palmitin. 

If hydnocarpic and chaulmoogric esters are specific for leprosy, 
Pangium edule oil would be an admirable remedy to use, since 
it would be easy to administer because of its low melting point 
and consequent fluidity, but would probably be slow in its action 
on account of the relatively small amounts of these acids pres- 
ent. On the other hand, Hydnocarpus alcalse oil would be much 
more difficult of administration because of its being a solid 
even at the ordinary temperature in Manila (30°C). 



The report of an investigation of the oils from Pangium edule 
and Hydnocarpus alcalse and of the cyanogenetic glucoside, gy- 
nocardin, is given. Their properties are discussed in some detail. 



REVIEW 

A Laboratory Guide | to the Study of j Qualitative Analysis | based upon the I 
Application of the Theory of | Electrolytic Dissociation | and the Law 
of Mass Action | by | E. H. S. Bailey, Ph. D. | professor of chemistry 
in the University of Kansas | and | Hamilton P. Cady, Ph. D. | pro- 
fessor of chemistry in the University of Kansas | Eighth Edition | 
revised by | Paul V. Faragher, Ph. D. | assistant professor of chemistry 
in the University of Kansas | in collaboration with the authors | 
Philadelphia | P. Blakiston's Son & Co. | 1012 Walnut street | Copy- 
right, 1916. Cloth pp. i-x+ 1-294. Price $1.50. 

The eighth edition of Professors Bailey and Cady's book, re- 
vised by Assistant Professor Faragher, is essentially the same as 
the seventh edition except for some minor modifications, con- 
sisting in the amplification of some sections difficult of com- 
prehension by the student. 

The scope and volume of this edition is, in general, the same 
as that of previous editions. The introductory part covers 24 
pages ; 126 pages are devoted to the reactions of the cations and 
a table for their systematic analysis, and 117 are taken up with 
the anions, also with a table for their systematic analysis. None 
of the rare metals are treated, except gold and platinum. A 
table for use in the examination of an unknown substance and 
another table of solubilities close the text. 

The introductory portion is furnished with a brief and accu- 
rate discussion of electrolytic dissociation and the mass action 
law as it is applied to qualitative analysis. The experiments 
given for the reactions of the individual ions are well selected; 
the tables for the separation of cations and anions into groups 
and the tests for their identification includes some good tests, 
for example, the dimethyl glyoxime reaction with nickel ions; 
while such subjects as hydrolysis, oxidation, and reduction are 
treated clearly in proper places in the book, thus rendering it 
easy for the student to comprehend their practical application. 

The mechanical details of the book are good and misprints 
are few (p. 68, 4H 4 instead of 4H 2 0). The type is clear, and 
the illustrated tables are well gotten up. 

On the whole, the book is a good, brief laboratory guide for 
students in qualitative analysis. 

F. PEftA. 









PUBLICATIONS FOR SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS— Continued 



A FLOKA OF MANILA 
By Elmer D. Merriu. 

Order No. 419. Paper, 490 panes, $2.50, 
postpaid. 

Practically a complete flora of the cul- 
tivated areas in the Philippines. Descrip- 
tions, with keys, t>f over 1,000 species, 590 
Otnera, and 136 families, with native names, 
glossary of technical terms, etc 



PHILIPPINE DIPTF.R0CARP FORESTS 
By WnjjlAAi H. BRown end Donald M. 

Matti: 

Order No. 432. Paper, 150 pages, 1 map, 
13 plates, and 12 diagrams, $1.00, 
postpaid. 
In Philippine Dlpterooarp Forests tha 
authors present a very comprehensive discus- 
sion of the growth and development of dip- 
terocarp trees and of the other elements of 
lowland Philippine forests. 



INDO-MALAYAN WOODS 
By Fred W. Foxwokihy 



ZOOLOGY— Continued 
A MANUAL OF PHILIPPINE BIRDS 

By Richard C. McGregor 



A Manual of Philippine Birds contains 
In oompact form descriptions of all the 
known speoles of Philippine birds. The usual 
keys and diagnoses of orders, families, and 
genera help the novice In identification. 



By David Starr Jordan and Robert Ea*l 
Richardson 



This list will be found a convenient guide 
to the synonymy of Philippine ichthyology. 
The nomenclature Is thoroughly revised, and 
the distribution of each species within the 
Philippine Islands Is given. 



In Indo-Malayan Woods, Doctor Fox- 
worthy has brjouoht together a Isrge amount 
of acourate information oonoerning treesi 
yielding woods of economic value. 



By W. Schultzb 



A LIST OF MAMMALS OF THE 
PHILIPPINE ISLANDS, EXCLU- 
SIVE OF THE CETACEA 

By Ned HolU3tek 

Order No. 41S. Paper, 64 pages, $0.50, 
postpaid. 
The dlstributii 



Paper, 19S pages, $1.00, 



This catalogue includes the names of all 
species of Coleoptera that have been recorded 
from a definite looality In the Philippine 
Islands. References to original descriptions 
and other Important notes are given. Tha 
economic appendix includes comment on 
those species of beetles which are known to 
be Injurious or beneficial to man. 



PRICES ARE IN UNITED STATES CURRENCY 

Orders for these publications may be sent to the BUSINESS MANAGER, 
PHILIPPINE JOURNAL OF SCIENCE, BUREAU OF SCIENCE, MANILA. P. I., 
or to any of the agents listed below. Please give order number. 



The Macmlllan Company, 64—66 Fifth Avenue, New York, V. S. A. 

Wm. Wesley & Son, 2 8 Essex Street, Strand, London, AV. C, En; 

Martinus Nijhoff, Lange Voorhaut 9, the Hague, Holland. 

Mayer & Mullet', Prlnz Louis Fordfnandstrasse 2, Berlin X. W-, Germany. 

Kelly & Walsh, Ltd., 32 Baffles Place, Singapore, Straits Settlements,. 

A. M. & J. Ferguson, 19 Bnillje Street, Colombo, Ceylon. 

Thaeker, Spink So Co., P. O. Box 54, Calcutta. India. 



BRILL, HARVEY C. 
HEISE, GEORGE W. The 

the normal constituents 
BRILL, HARVEY C. A chemica 

Pangium edule and of Hydnc 
Review 



The "Philippine Journal of Selenco" is issnecl as foil 
Section A. Chemical and Geological Sciences and the Inc 

Section B. Tropical Medicine 

Section C. Botany 

Section D. General Biology, Ethnology, and Anthropology 

tion B began with Volume V) 

Entire Journal, Volume II, III, IV, or V 

Entire Journal, beginning with Volume VI — 

Single numbers ( except of Volume I ) 

"Each section is separately paged and indexed. 
Authors receive lOO copies of their papers free. 
Volume I, 1906 (not divided into sections) and supplement, sold 

only with » complete file of section A, B, or C 10.00 

Supplement to Volume I (botany) - 3.50 

Volume I (without supplement) , sold only with a complete file of 

section A, B, or C 6.50 

Single numbers of Volume I - ...- - .75 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, P. I., or to any of the agents 
listed below: 

AGENTS 

The Macmillan Company, 64—66 Fifth Avenue, New York City, V. S. A. 

Win. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinus NijhoiY, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdlnandstrasse 2, Berlin, N. W., Germany. 

Kelly & Walsh, Limited, 32 Baffles Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 19 Baillie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



Entered at the post office at Manila, P. I., at tecond-cla»» matter. 









VOL. XII, SEC. A, No. 2 



Makch, 1917 



THE PHILIPPINE 

JOURNAL OF SCIENCE 



ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D.; J. R. WRIGHT, Ph. D.; G. W. HEISE, M. S. 

J. C. WITT, Ph. D.; T. DAR JUAN, A. B.; A. H. WELLS, A. B. 

R. C. MCGREGOR, A. B.; H. E. KUPFER, A. B. 




MANILA 
BUREAU OP PRINTING 

1917 






ETHNOLOGY 

A VOCABULARY OF THE IGOROT LAN- 

GUAGE AS SPOKEN BY THE 

BONTOC IBOROTS 

By Walter Clayton Clapp 

Order No. 408. Paper, 89 pages, $0.75. 

postpaid. 
The vocabulary is given In Igorot-English 
and English-lgorpt, 



THE NABALOI DIALECT 

By Otto Schekber 

and 

THE BATAXS OE PALAWAN 

By Edward Y. Miller 

Order No. 403. Paper, $0.25; half mo- 

rocco, $0.75; postpaid. 

The Nabaloi Dialeot (65 pages, 29 

plates) and the Bataks of Palawan (7 

Paget, 6 plates) are bound under one cover. 



THE BATAN DIALECT AS A MEMBER 

OF THE PHILIPPINE GROUP 

OP LANGUAGES 

By Otto Scueerer 

and 

"F" AND "V" IN PHILIPPINE 

LANGUAGES 

By Carlos Everett Conant 

Order No. 407. 

These two papers are Issued under one 
cover, 141 pages, paper, $0.80, postpaid. 




PUBLICATIONS FOR SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



ETHNOLOGY— Continued 



By Najeeb M. Saleeby 

Order No. 405. Paper, 107 pages, 16 

plates, 5 diagrams, $0.25; half mo- 
rocoo, $0.75; postpaid. 

This volume deals with the earliest 
written records of the Moros In Mindanao. 
The names of the rulers of Maglndanao are 
recorded In five folding diagrams. 



NEGRITOS OF ZAMBALES 

By William. Allan Beeb 

Order No. 402. Paper, 83 pages, 62 

plates, $0.25; half moroooo, $0.75; 
postpaid. 
Plates from photographs, many of whioh 
were taken for this publication, show orna- 
ments, houses, men making fire with bamboo, 
bows and arrows, dances, and various type* 
of the people themselves. 



INDUSTRIES 

PHILIPPINE HATS 

By C. B. Robinson 

Order No. 415. Paper, 66 pages, 8 
plates, $0.50 postpaid. 

This paper Is a oonoise record of the 
history and present condition of hat making 
In the Philippine Islands. 



THE SUBANUNS OF SINDANGAN BAT 

By Emerson B. Christie 

Order No. 410. Paper, 121 pages, 1 
map, 29 plates, $1.25, postpaid. 

Sindangan Bay is situated on the north- 
ern coast of Zamboanga Peninsula. The Su- 
banuns of this region were studied by Mr. 
Christie during two perfods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanuns at 
work and at play; their Industries, houses, 
altars, and Implements; apd the people 
themselves. 



By Herbert S. Walker 

Order No. 412. Paper, 145 pages, 10 
plates, 1 map, $1.25, postpaid. 

Considered from the viewpoint of prac- 
tical utility, Mr. Walker's Sugar Industry 
In the Island of Negro? is one of the most 
Important papers published by the Bureau 
of Science. This volume Is a real contribu- 
tion to the subject; it is not a mere com- 
pilation, for the author was in the field and 
understands the conditions of which he 
wri tes. 



THE HISTORY OF SULU 

By Najeeb M. Saleeby 

Order No. 406, Paper, 275 pages, 4 

maps, 2 diagrams, $0.75, postpaid. 
In the preparation ef his manuscript for 
The History of Sulu, Doctor Saleeby spent 
muoh time and effort in gaining access 
to documents in the possession of the Sultan 
of Sulu. This book is a history of the 
Moros in the Philippines from the earliest 
times to the American occupation. 



By Charles S. Banks 

Order No. 413. v Paper, 53 pages, 20 
plates, $0.75. postpaid. 
In A Manual of Philippine Silk Culture 
are presented the results of several years' 
actual work with silk-producing larvas to- 
gether with a description of the new Philip- 
pine race. 



THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 

Vol. XII MARCH, 1917 No. 2 

THE STUDY OF COPRA AND OTHER COCONUT PRODUCTS » 

By Alvin J. Cox 
(From the Bureau of Science, Manila) 

For several years the chemists, assisted by the botanists, of 
the Bureau of Science have been employed, as time permitted, 
in a study of the products of the coconut. Many investigations 
have been made and published by the Bureau of Science on the 
subject. 2 Coconut oil as such and in the unpressed copra ranks 
among the three most important exports from the Philippine 
Islands and, therefore, warrants careful study. There are two 
large operating vegetable oil plants in the Philippine Islands, 

1 Received for publication February, 1917. 

J Freer, Paul C, On the water relations of the coconut palm (Cocos 
nucifera) — On the oil produced from the nuts — The factors entering into the 
rancidity of the oil — The insects attacking the trees — Introduction, This 
Journal (1906), 1, 3-5. Copeland, E. B., On the water relations of the 
coconut palm (Cocos nucifera), ibid. (1906), 1, 6-57. Walker, Herbert S., 
The coconut and its relation to the production of coconut oil, ibid. (1906), 1, 
58-82. Walker, Herbert S., The keeping qualities of coconut oil and the 
causes of its rancidity, ibid. (1906), 1, 117-142. Banks, Charles S., The 
principal insects attacking the coconut palm, ibid. (1906), 1, 143-168, 
211-228. Richmond, George F., Purification of coconut oil, ibid., Sec. A 
(1908), 3, 45-47. Walker, Herbert S., Notes on the sprouting coconut, on 
copra, and on coconut oil, ibid., Sec. A (1908), 3, 111-135. Gibbs, H. D., 
and Agcaoili, F., On the detection and determination of coconut oil, ibid., 
Sec. A (1908), 3, 371-375. Pratt, David S., Copra spoilage on a large scale, 
ibid., Sec. A (1913), 8, 439-441. Pratt, David S., The coconut and its 
products with special reference to Ceylon, ibid., Sec. A (1914), 9, 177-199. 
Brill, Harvey C, Parker, Harrison O., and Yates, Harry S., Copra and 
coconut oil, ibid., Sec. A (1917), 12, this number. Parker, Harrison O., and 
Brill, Harvey C, Methods for the production of pure coconut oil, ibid., Sec. A 
(1917), 12, this number. 

149522 49 




50 The Philippine Journal of Science v»vt 

and several small ones are now being erected or planned. The 
work of the Bureau of Science has been done from time to time 
in order to settle some special problem, and all of it has been 
beneficial in pointing the way to more extended problems. The 
studies by Walker 8 on the keeping qualities of coconut oil and 
the causes of its rancidity and notes on the sprouting coconut, 
on copra, and on coconut oil have been very helpful to the public, 
and his papers have been in great demand. For several years 
we have endeavored to find time to continue and extend this 
work and to accumulate information with regard to the com- 
position of the coconut, the hydrolysis and consequent destruc- 
tion of fat, the methods of drying, the methods for the most 
effective recovery of the oil, the. methods of analysis, the detec- 
tion of adulterants of coconut oil and of coconut oil as an 
adulterant of other oils, and the reduction to a minimum of 
the loss through deterioration of copra and coconut oil during 
transportation. Data showing the composition of many Phil- 
ippine soils have been published, and some of these concern 
the soils from the best coconut areas. 

In recent years I have been impressed by the large loss of 
coconut oil through spoilage and the unearned revenue that 
the Philippine Islands would secure if all the copra produced 
were of a high grade. Furthermore there is a loss in shipping 
poorly cured copra, not only in the deterioration due to mold 
and bacterial action, but also in the transportation of the excess 
moisture. 

The work begun by Dr. Paul C. Freer and by Mr. H. S. Walker 
has been continued by Dr. Harvey C. Brill and Mr. Harrison 
0. Parker, the botanical work being done by Dr. Harry S. Yates. 

In California large quantities of deciduous fruits are opened, 
treated with sulphur dioxide (the fumes of burning sulphur) 
to protect them from bacterial or mold action, and subsequently 
dried. The action of sulphur dioxide is to kill all mold spores 
and to soften the cell walls of the fruit so that drying is facili- 
tated. Enough of the sulphur dioxide remains in the meat to 
prevent the growth of new mold spores during drying, if drying is 
completed in a week or two. 

We have successfully applied this method to the drying of 
coconut meat. Coconuts opened and treated with the fumes 
of burning sulphur at the Bureau of Science during a severe 
rainstorm, which subsequently received no artificial drying or 
exposure to the sunshine, remained perfectly white for a period 

' Loc. cit. 



xii. a, 2 Cox: Study of Copra 51 

of two weeks. Copra on hand after many months is still of 
excellent quality. The box in which the treatment with sulphur 
dioxide is made must be fairly tight, but not air-tight. There 
must be circulation enough to keep the sulphur burning. In the 
tapahan (Filipino grill for drying coconut meat) method of dry- 
ing copra the coconut meat frequently begins to mold before 
the drying is begun, and before the drying has proceeded far 
enough to inhibit the growth of mold, considerable deterioration 
has taken place. In the sulphur process the nuts can be sub- 
jected to sulphur dioxide before mold has started to grow. With 
proper organization and routing of the work, the labor cost when 
the sulphur dioxide method is used will not exceed that in the 
tapahan. 

Compared with the tapahan method the sulphur dioxide pro- 
cess is exceptionally clean : the copra is preserved and bleached 
by the sulphur dioxide and yields very white copra ; there is no 
loss of oil during the treatment or during the drying; an ex- 
ceedingly uniform copra is obtained, and its keeping quality is 
improved; and the oil expressed from the copra is practically 
colorless, is free from rancidity, is pronounced equal to, or better 
than, the best Cochin oil, and usually will sell for at least 10 
centavos 4 a kilogram (2 cents or more a pound) more than 
ordinary oil. At 10 centavos a kilogram there is a difference 
of about 4 pesos per 63.25 kilograms (1 picul) of copra. Storage 
conditions and the problem of storage must receive careful con- 
sideration even with first-class copra ; however, this will become 
less and less of a problem as more and more of the Philippine 
copra is consumed in local oil mills. 

Other experiments have been in progress in an effort entirely 
to eliminate the drying process and to extract the oil from fresh 
coconut meat. 

Lack of means for producing a good grade of copra or oil 
from the fresh nut has not been the only obstacle in the way 
of the improvement of the coconut industry. Most dealers have 
been contented with a poor copra and could not see any advantage 
to themselves in being able to secure a better product with a 
higher oil content. Dealers say that it has been their custom 
to put good copra with the poor, but the vegetable-oil companies 
will buy the good copra at a premium as soon as they can get 
enough to run a mill for a day or two once or twice a month. 
Many dealers now realize that properly dried copra is worth 

4 One peso Philippine currency equals 100 centavos, equals 50 cents United 
States currency. 



52 The Philippine Journal of Science i»" 

more — not only in proportion to the reduced water content, but 
also on account of its improved keeping qualities and the higher 
grade and value of the oil that can be obtained from it. If 
necessary, it will be productive of great results for the Govern- 
ment to penalize smoked, colored, dirty, moldy, or imperfectly 
dried copra and to subsidize the higher grades of copra until 
dealers become aroused and demand them, which they certainly 
will do in time. Such action will not only protect the consumer, 
but will increase the revenue of the producer. In order to estab- 
lish definite grades and a certain market, a satisfactory system 
of classification and standards must be devised, as the Bureau of 
Science has long been advocating. At present no definite grades 
for a given region exist, owing to the unwillingness of the inhabit- 
ants properly to dry the copra. In the Manila markets the terms 
Cebu sundried, fair marketable Manila, and Laguna are known, 
although ill defined. Cebu sundried usually commands about 75 
centavos per 63.25 kilograms (1 picul) more than Laguna. A 
step toward the solution of the question of standardization has 
been made by the Visayan Refining Company. Nine months of 
experience gave this company 65.5 per cent oil in the Cebu sun- 
dried copra used in its mill. The company has taken this as 
the average for a good grade of copra in the Cebu market, and 
for such copra it pays the market price and, in addition, when 
the copra is white, guarantees the consignee premiums for addi- 
tional oil content as follows : 



66 12.5 

67 25. 

68 37.5 

By this system the producer cannot lose, and the possibility 
of receiving a bonus is an incentive to dry properly and to pro- 
duce a higher grade copra. 

In order satisfactorily to establish grades of copra, certain 
facts must be considered. Well-prepared copra is white, but 
the discoloration of black copra may be due to its being smoked 
when dried on the tapahan or to mold action. Copra is either 
wet or dry. There are various degrees of wetness. The follow- 
ing paper indicates the permissible amount of water if copra 
is expected to be mold-free in storage. The character of copra 
depends to a certain extent upon the variety of the nut and the 
region in which it is grown; therefore it is probable that it will 
be necessary to establish regional grades of copra. It is my 
opinion that on the basis of oil content in relation to moisture 



xii.a.2 Cox: Study of Copra 53 

content, cleanliness and freedom from dust, foreign matter or 
adulteration, free fatty acids, freedom from unripe nuts, and 
color and general appearance uniform rational grades for copra 
can be established throughout the Philippine Islands. The use 
of green nuts for copra in Samoa is prohibited. There is a law 
which provides that nuts for making copra must be allowed to 
drop from the trees. There are varieties of Philippine coconuts 
whose nuts do not drop when ripe. Oil content alone does not 
define a good copra, as the oil content of badly molded copra 
may be on a par with that of properly prepared copra, owing 
to the simultaneous destruction by molds of oil and of tissue in 
relative proportions. Also the oil content of the coconuts may 
vary with the region in which they are grown or with the variety 
of the nut. The enforcement of grading should be easy, for 
the color and appearance can be readily determined by inspec- 
tion, and an inspector should soon become sufficiently expert tc 
judge the quantity of moisture by feel. There are also several 
simple tests that may be applied to substantiate his judgment. 
If copra is excessively wet, it is usually hot ; if sufficiently dried. 
a piece of copra breaks when the two edges are pressed together. 
The width of the dark line showing on the fresh broken edge 
is a criterion of the moisture content. An expert can also judge 
of the oil and water content by lighting a sliver of copra. If 
it burns readily, the copra is fairly dry and has a high oil content ; 
if it is moderately wet, the flame will sputter as the moisture 
is evaporated; if very wet, it will not burn. 

The studies that follow contribute to our knowledge of copra 
and coconut oil and outline the means of obtaining a less acid 
and less rancid oil; this will command a higher market price 
and will bring the Philippines to the front for the quality of 
its copra and its coconut oil; this, in turn, will increase the 
revenue from Philippine coconut plantations. 



COPRA AND COCONUT OIL 1 

By Harvey C. Brill, Harrison 0. Parker, and Harry S. Yates 

(From the Laboratory of Organic Chemistry, Bureau of 
Science, Manila) 

Increasing demands for fatty foodstuffs have developed the 
vegetable-butter industry in the United States — an industry for- 
merly restricted to England, Holland, and France. Refineries 
and hydrogenation plants have been able to purify and harden 
practically all of the animal and vegetable oils suitable for edible 
purposes into either simple or compounded vegetable lards and 
butters. The oil obtained from the coconut ranks high in im- 
portance among these. However, an enormous amount of co- 
conut oil is at present being used in the soap and glycerin in- 
dustries, rather than for the manufacture of edible products, 
because of the poor quality of the commercial oil and its conse- 
quent cheapness. 

The production of copra constitutes one of the leading indus- 
tries of many tropical countries. In the Philippine Islands prac- 
tically the entire annual crop of about 431,387,000 nuts, 2 with the 
exception of those used for local consumption, is turned into 
copra. Copra exports from the Philippine Islands for 1916 were 
72,277,164 kilograms, and oil exports were 16,091,169 kilograms. 3 

The annual exports represent approximately one third of the 
world's output of copra, 4 most of which finds its way to the mar- 

1 Received for publication February, 1917. The isolation and study of 
the molds growing on Philippine copra were made by Harry S. Yates, of 
the section of botany, Bureau of Science. 

' Cox, Alvin J., Bureau of Science Press Bull. (1916), No. 54. Computed 
from the yield of copra and based on the experiments of the Bureau of 
Science, which show that for the Philippine Islands 1,000 nuts yield about 
270 kilograms of copra. 

* Information furnished by the Insular Collector of Customs April 17, 
1917. 

* Smith, H. H., Coconuts. The Consols of the East. 2d ed. Tropical 
Life, London (1913), 362. Lewkowitsch, Chemical Technology and Ana- 
lysis of Oils, Fats and Waxes. Macmillan & Co. Limited, London (1914), 
22, 635. 

65 



56 The Philippine Journal of Science im 

kets of Great Britain, France, Holland, and the United States. 
There are but three modern coconut oil mills in operation in 
the Philippine Islands. These, together with the native mills, 
exported oil to the value of 5,641,003 pesos 5 in 1915. 8 

It is surprising to note that Philippine copra is quoted the 
lowest on the world's market. By calculating the value of Ceylon 
copra exported from 1908 to 1911 and comparing the figures with 
the value received for Philippine copra during the same period, it 
is found that the annual difference between what the Philippine 
Islands received and what they should have received in 1911 7 is 
more than 4,000,00 pesos and for the previous five years is more 
than 15,000,000 pesos. 

The reason for the low price is found in the poor quality of 
copra produced. Some attention has been paid of late years to 
the improvement of the quality, but concerning the cause of 
this inferiority only a little published data is available. Oil ob- 
tained from the usual Philippine copra is discolored and rancid 
and contains free fatty acids varying from 5 to 20 per cent 
(oleic acid). These conditions favor even further deterioration 
of the oil. 8 The quality of the oil depends primarily upon the 
condition of the copra at the time it is milled, and poorly pre- 
pared copra deteriorates rapidly with loss of oil and impairment 
of its quality. The poor quality of the copra is due to insufficient 
drying and unclean methods used in its production. It is the 
purpose of this paper to bring out analytical and botanical data 
relative to losses in copra and oil due to the faulty production of 
copra and to suggest means for improvement. Unless coconut 
meat is dried, immediately after opening the nuts, to a moisture 
content of approximately 6 per cent, it is attacked by various 
microorganisms, which causes a loss in oil content. The extent 
of the loss depends upon the length of time the meat retains 
sufficient moisture for mold growth. 

It was determined by experiment and observation that molds 
grow most luxuriantly upon copra with a moisture content of 
10 per cent or greater, which, as shown in Table I, is common in 
commercial copra. 

8 One peso Philippine currency equals 100 centavos, equals 50 cents 
United States currency. 

"Annual Rep. P. I. Bur. Customs (1915), 17, 18. 
'Pratt, D. S., This Journal, Sec. A (1914), 9, 186. 
"Walker, H. S., This Journal (1906), 1, 141. 



xii. a. 2 Brill, Parker, and Yates: Copra and Coconut Oil 



57 



Table I. — Moisture content of copra samples from various localities in the 
Philippine Islands. 



Locality. 


Water. 


Maxi- 
mum. 


Mini- 
mum. 




Per cent. 
29.1 
23.1 
22.2 
20.7 
24.7 
14.1 


Per cent. 
18.8 
14. 6 
17.fi 
14.4 | 
15.5 
10.4 | 















It was deemed essential to determine the loss in weight of 
commercial copra stored and handled under ordinary conditions, 
especially as laboratory experiments conducted on copra under 
the most favorable moisture conditions for mold growth (10 to 
20 per cent) showed a loss of 25 per cent in total oil content. 
Table III gives losses for copra stored in bodegas. The data 
in Table II were obtained by analysis of partially dried copra 
stored in a container at room temperature for fifteen days. The 
moisture conditions were regulated throughout the experiment 
to favor maximum mold growth, that is, from 10-20 per cent 
water. The loss represents the combined action of green, brown, 
black, and white molds. 



Table II. — Effect of the four common molds — green, brown, black, and 
white — growing together on copra. 



Series No. 


Weight 
of copra. 


Oil in 
copra 

before 
mold 

action. 


Oil after 
mold 
action. 


Lobs of oil. 


Acidity 
as oleic 
acid after 

mold 
action." 




a. 

17.10 

17.50 

16.62 

15.96 

16.14 

84.20 


g- 

5. 0838 
5. 2027 
4. »411 
4.3571 
4. 7984 
25.0326 


g. 

3. 8138 
3.8882 
3. 7369 
3. 3167 
3.4468 
18. 6020 


O. 

1.2700 
1. 3145 
1.2042 
1.0404 
1.3516 
6.4306 


Per cent. 
24.9 
25.2 
24.3 
23.8 
28.1 
25.6 


Per cent. 
15.6 
12.4 
13.0 
9.0 
11.4 
10.0 















« Acidity of oil expressed from fresh coconut meat is always about 0.2 per cent calculated 
oleic acid. 



58 



The Philippine Journal of Science 



It is obvious that if there is a loss in oil there must also be 
a loss in weight of the copra from this source. Buyers of copra 
recognize that a loss in weight occurs when copra is stored, and 
they are governed in their purchase by this factor. The loss 
in weight when commercial copra is stored is shown in Table III. 

Table III. — Loss in weight of copra from various provinces when stored 
in commercial bodegas. 



Place of production. 



Albay Province: 

Legaspi 

Do 

Do 

Camarinee Province: 

Daet 

Do 

Do... 

Do. 

Capiz Province: 

Calivo 

Do 

Do 

Jolo - 

Laguna Province: 
PagBanjan 

Do 

Do 

Do 

Do 

Do 

Do. 

San Pablo.. 

Do._ 

Do 

Pangasinan Province: 

Dagupan 

San Carlos 

Romblon 

Do 

Do 

Magallanet 

Do 

Samar Province: 

Catarman 

Catbalogan 

Sorsogon Province: 
Bulan 

Do... 

Gubat.. 

Do 

Do 

Do._ 

Surigao Province 



Loss in 
weight. 



Per cent. 
9.61 



0.00 
6.25 
2.5 



8.77 
10.07 
4.5 
1.5 

16.60 
8.41 
2.0 

10.25 
6.5 

12.25 
7.00 

10.48 

12.71 



8.77 
10.00 

3.87 

7.5 
11.5 

2.86 



0.86 
2.5 



1.5 
0.91 
10.22 
6.38 
8.15 
6.5 



xii. a, * Brill, Parker, and Yates: Copra and Coconut Oil 59 

Table III. — Loss in iveight of copra, etc. — Continued. 



Place of production. 


Days 

stored. 


Loss in 
weight. 


Tayabas Province: 


25 
30 
28 
26 
19 
20 
29 
28 
18 
18 
31 
18 
12 
14 
24 
23 
29 
15 
25 


Per cent. 

1.00 
0.4 

17.74 

12.97 
1.76 

12.07 

14.68 
2.90 
1.75 
6.53 

18.69 
9.09 

14.05 
9.75 
8.00 

10.00 
7.50 
8.06 

12.5 












Do- - --- 








Do ■ _ 


Do.. _ 


Do ._ 




Do _ 


Do 











There is a further loss in weight of copra on shipboard, which 
the copra dealers estimate at from 3 to 6 per cent, depending 
primarily upon the length of time of storage before shipment. 
With the high freight rates that have prevailed for the past year 
for copra shipments to the United States, the loss due to paying 
freight on loss of weight alone is considerable ; even under normal 
conditions this loss is too large to be ignored. 

Freight rates at 50 pesos per ton with loss in weight of 6 
per cent -amount to 3 pesos for each ton of copra. When the 
product is not completely dried, there results not only an unneces- 
sary expense of handling an excessive amount of water, but also 
the conditions are most favorable to mold growth with conse- 
quent loss in quality and quantity of the oil. It is the producer 
who suffers, 9 for the purchaser reduces the price to cover not 
only the extra water, but also the extra handling and transporta- 

* While the price paid to the producer is less for his poor product than 
it would be for a higher grade product, some question exists as to his 
suffering any excessive monetary loss. Consideration should be given to 
the fact that he has no great amount of money invested in apparatus, that 
he can employ unskilled labor, and that he expends no great amount of 
care in the preparation of his product. The difference in the cost of 
producing poor and of producing good copra is hardly made up by the 
present discriminating price. The production of poor copra should rather 
be considered an economic loss to the Philippine Islands. 



60 The Philippine Journal of Science vm 

tion expenses. That these losses incident to storage and ship- 
ment extending over periods of from two to four months are not 
due entirely to the evaporation of water is evident from Table IV; 
showing temperature, carbon dioxide, and weight loss relation- 
ship of stored Laguna copra. It is obvious from these figures, 
which show an increased temperature with a corresponding in- 
crease in carbon dioxide over the normal atmosphere condition, 
that slow combustion is taking place with the formation of carbon 
dioxide and water, necessarily at the expense of the meat and 
the oil. 

The increase in temperature of the copra parallels the in- 
crease in carbon dioxide content of the atmosphere surrounding 
the copra, rises to a maximum at the same time, and decreases 
along with the decrease of the carbon dioxide concentration, 
proving that the heat is produced by the combustion of the copra. 



Table IV. — Temperature and carbon dioxide relationship of commercial, 
stored copra. Temperature of atmosphere, 29° C. 



Copra. 


Date. 


Tem- 
per- 
ature. 


Carbon 
dioxide 
(CO2) in 
atmos- 
phere over 
stored co- 
pra. 






°C. 


Per cent. 




( September 20 


40 


0.4 


Lot I _ - 


September 23 
September 25 


40 
38 


0.4 
0.3 






September 26 


38 


0. 2-0. 3 




/ September 26 


50 


1.0 




September 27 


55 


1.6 




September 28 


65 


1.4 




September 29 


53 


1.2 




September 30 


50 


1.0 




October 2 


48 


0.8 


Lot II - - 


October 4 
October 6 


48 
45 


0.8 
0.6 






October 7 


41 


0.4 




October 9 


40 


0.4 




October 10 


38 


0.4 




October 11 


38 


0.2 




October 12 


36 


0.2 




October 13 


35 


0.2 




September 26 


45 


0.8 




September 27 


44 


0.8 


j 


September 28 


42 


0.6 


Lot III - : 


September 29 
September 30 


40 
40 


0.6 

0.4 


| 




October 2 


38 


0.4 




October 4 


38 


0.2 




October 6 


36 


0.2 



xii, a, 2 Brill, Parker, and Yates: Copra and Coconut Oil 



61 



The data in Table IV was obtained by measuring the average 
temperature of a lot of copra with a moisture content correspond- 
ing to the commercial samples described in Table I. The quantity- 
amounted to over 200 piculs. The carbon dioxide measurement 
was made by withdrawing air from the center of the pile and 
determining the percentage of carbon dioxide by means of a 
special absorption apparatus. The average temperature of the 
air outside was 29° C, and the percentage of carbon dioxide in 
the air was always less than 0.10. 

Five bags belonging to lot II were weighed before and after 
storage in the general pile from which the carbon dioxide and 
temperature determinations were made. 



Table V. — Weight loss in commercial copra stored twenty-five days 
bodeoas. 



in 



1 

Bag No. 


Weight. 


Septem- 
ber 26. 


October 
19. 


Loss. 




Kilos. 


Kilos. 


Per cent. 


i 


45 
56 
73.5 
70.5 


40.5 10.0 
48. 5 j 13. 4 
66.5 I 9.5 


8 






"" " """ " 







The copra in the general pile after storage was badly damaged, 
particularly the bags of the lower layers, which were matted 
together into a solid mass of black, foul-smelling, decayed ma- 
terial. It is not reasonable to presume that in a large pile of 
sacked copra much moisture will escape of its own accord. How- 
ever, so long as moisture is sufficient in amount to cause mold 
growth, the combustion brought about within the pile, as was 
most evident during the periods of highest temperature condi- 
tions, causes liberation of the moisture. The cause for a decrease 
from a maximum temperature and from carbon dioxide condi- 
tions to normal is influenced by two factors — there is a loss in 
water to a point where the mold will not flourish, and the tem- 
perature reached by the copra is sufficient to inhibit if not en- 
tirely to destroy the mold spores. 10 

10 Rather, J. B., Journ. Ind. & Eng. Chem. (1916), 8, 604, has shown that 
cottonseed containing a percentage of moisture greater than a certain 
maximum heats when piled and that this heating continues until the 
moisture content is lowered. An increase in the free fatty acids and a 
darkening of the oil are results of the heating. 



62 



The Philippine Journal of Science 



Well-dried copra neither became hot nor evolved carbon dioxide 
when piled in storage, demonstrating that this combustion is 
due to the presence of higher quantities of moisture together with 
the consequent action of the mold. 

Various factors enter into the analysis of copra from the point 
of view of oil loss. Curiously, only in exceptionally bad copra 
is the percentage of oil found materially lower than would be 
expected in good copra. This is due to the fact that both meat 
and oil are attacked by the microorganisms lowering the total 
weight of the sample. Therefore it is necessary that the total 
weight of the oil in the original piece of fresh meat be known 
in order that the percentage of total loss may be calculated 
from the actual weight of oil before and after deterioration. 

Table VI shows the apparent loss of oil calculated from the 
percentages and the true loss in oil for the same piece of copra 
and clearly illustrates our statement above. 

Table VI. — Oil losses in copra calculated on oil percentages and on total 
oil content. 



Dry meat before mold action, calcula- 
ted. grams. 

Dry meat after mold action do... 

Oil in dry meat before mold action. .per cent. 
Oil in dry meat after mold action, calcula- 
ted per cent- 
Total oil before mold action, calcula- 
ted grams. 

Total oil after mold action do 

Apparent loss of oil from percent- 
ages - per cent- 
True loss of oil do... 



9. 1645 
6. 3465 



11.7834 
10. 2104 
67.5 



.951.1 
. 6846 



16. 1955 


13.3213 


14. 4030 


11.9880 


66.8 


66.8 


47.1 


46.7 


10.8156 


8.8744 


6.7880 


5.6010 


19.7 


20.1 


37.2 


36.8 



The above results conclusively show that the true loss of oil 
is very much greater than the apparent. The apparent change 
does not take into account the weight of the fiber destroyed; 
consequently the apparent loss is always smaller than the true loss 
and in certain instances might appear negligible or that there 
had been a synthesis of oil by mold growth. Indeed Walker, 11 in 
some experiments where he did not take this factor into consid- 
eration, obtained results which indicate a synthesis of oil by 
microorganisms. Such results show the worthlessness of data 



Op. cit., 117. 



xii. a. 2 Brill, Parker, and Yates: Copra and Coconut Oil 63 

calculated on the basis of the molded sample for demonstrating 
the quantity of oil actually decomposed. 

MICROORGANISMS AND THEIR EFFECT ON COPRA AND COCONUT OIL 

Comparatively little attention has been given to the effect 
of microorganisms on coconut oil. Walker 12 has called attention 
to the fact that they do seriously affect the quantity and quality 
of the oil in copra. However, his experiments were few and 
devised primarily to show the action of the organisms on the 
oil irrespective of the amount of moisture present in the copra. 
When fresh coconut meat is exposed to the air, various fungi 
make their appearance, and as the moisture content of the meat 
becomes less, these fungi are succeeded by others. Certain fungi 
always appeared on the fresh meat and others on fairly dry 
copra, and so it seemed probable that each of the fungi con- 
cerned was restricted more or less closely to a definite condition 
of moisture in the copra. The life history of each of the more 
important organisms was studied to determine whether the effect 
on the oil caused by the different species of fungus might not 
vary. Special attention was paid to the rate of growth, the 
moisture requirements, and the effect upon the oil. 

Our early observations confirmed Walker's 13 statement that 
two classes of plants known as bacteria and fungi are present and 
may be concerned in the deterioration of copra. Very early in 
the investigation we determined that bacteria normally play 
only a very minor role in the deterioration of copra, since a 
moisture content sufficiently high to favor bacterial growth is 
not found in ordinary copra. Therefore the investigation has 
not been concerned with the action of bacteria beyond a general 
confirmation of Walker's conclusions that bacteria grow only on 
fresh coconut meat or on copra, whose moisture is very high. 
Furthermore they cause little if any loss of oil even under con- 
ditions most favorable for their growth. Bacteria do seriously 
affect the appearance of the copra and the quality of the oil, as 
by their action they break the copra down into a slimy mass with 
an offensive odor. However, the fungi play a much more im- 
portant part in the deterioration of the oil in copra than do 
bacteria. 

Copra containing less than 20 per cent of water is practically 
free from bacterial action, and even above this moisture content 
the deterioration caused by fungi is of far greater importance. 
Not only are the moisture requirements of most of the fungi 

"Op. cit., 117. "Op. cit. (1906), 1, 135. 



64 The Philippine Journal of Science i«n 

with which we are concerned lower than those of the bacteria, 
but the fungi also cause a very considerable loss of oil and se- 
riously impair its quality. They likewise injure the appearance 
of the product and help to collect dirt and other foreign material. 
If mold action is prevented, the action of bacteria is prevented 
at the same time and by the same methods. 

The molds occurring upon copra belong to several groups of 
fungi, but they closely resemble each other in their vegetative 
habits and in their effect. 14 

We have found four molds constantly occurring upon moldy 
copra and coconut meat. The spore masses of these four molds 
differ greatly in color, and hence the molds are very readily 
distinguished even without the use of a microscope. In the or- 
der of the moisture necessary for their growth these molds are 
Rhizopus sp. (white mold), a mold occurring only upon fresh 
meat and there forming loose masses of white mycelium with 
many black sporangia; Aspergillus niger Van Tiegh. (black 
mold) , a mold occurring on copra with a relatively high moisture 
content and producing black spore bodies which give the mold a 
black color; Aspergillus flavus Link (brown or yellow mold), a 
mold occurring most commonly on moldy copra. The spore 
masses are first greenish yellow, later turning brown; PeniciU 
lium glaucum Link (green mold), a mold producing green spore 
masses and common on copra, especially that containing a low 
percentage of moisture. 

RHIZOPUS SP. (WHITE MOLD) 

Rhizopus occurs only upon fresh coconut meat and then only 
when the surrounding air is in a practically saturated condition. 
When moisture conditions are favorable, the growth of this mold 
is very luxuriant. It spreads by means of stolons, and in from 
thirty-six to forty-eight hours the mycelium frequently entirely 

14 Molds grow from extremely minute single-celled bodies called spores, 
which correspond to seeds in the higher plants. When a spore of one of 
these fungi germinates, it puts forth one or more delicate, colorless filaments, 
which grow in length, branch repeatedly, and work their way through the 
coconut meat or the copra. At the time this thread is growing, it is 
breaking down the oil and cell walls of the coconut by means of enzymes, is 
using the material to build up its own tissue, and is liberating in the process 
carbon dioxide and water. When the fungus fruits or produces spores, the 
plant itself becomes visible to the observer. We should remember when 
considering mold action that the part active in the destruction of the oil is 
invisible and that it is active in this destruction almost from the moment 
the spore germinates, which takes place some time before the fungus 
becomes visible. 



xii, a. 2 Brill, Parker, and Yates: Copra and Coconut Oil 



65 



covers a piece of coconut meat 10 centimeters in diameter with 
a tangled mass of aerial mycelium, which may attain a height 
of from 3 to 5 centimeters and which is specked with small black 
sporangia. The aerial mycelium collapses upon the slightest 
drying. The spores germinate in about six hours in a hanging 
drop of coconut decoction. Although this mold grows rapidly 
and destroys a high percentage of the oil in the meat, it is prob- 
ably the least important of the four molds considered in this 
paper. It grows only upon fresh meat, and hence its growth 
is checked and the plant killed almost as soon as drying com- 
mences. Rhizopus can rarely make any considerable growth, 
since the meat is usually treated or placed on the grate and heat 
applied within the period necessary for the germination of the 
spore. 

Morphology. — Mycelium hyaline, young hyphse clear, older 
ones filled with granular material. Filaments about 8 fx in dia- 
meter. Sporophores arising in groups of two or three; stalk 
about 0.8 to 1 millimeter in height, 10 /x in diameter. Sporan- 
gium 100 to 140 fi in diameter, first globular, the wall later rup- 
turing and folding back. Spores black in mass. The single 
spore brownish, smooth, 7 to 9 by 4 to 5 ft. 

Oil loss caused by Rhizopus sp. {white mold). — Table VII 
shows the oil loss due to white mold acting upon grated coconut 
meat in a saturated atmosphere during a period of about ten 
days. The experiment was performed and the oil loss calculated 
as follows: Freshly grated meat from one coconut was divided 
into several samples of approximately 10 grams each and ac- 
curately weighed into separate tared extraction thimbles. The 
original total oil of the several samples was calculated from 
the average analysis of two of the samples, and the remaining 
tubes and contents were sterilized and inoculated with spores 
of Rhizopus sp. After the mold had been allowed to act for a 
period of ten days, the thimbles were extracted with chloroform, 
and the loss in oil was calculated from the original total oil 
content and the total oil remaining after mold action. 

Table VII. — Oil loss due to the action of Rhizopus {white mold). 



Sample No. 




Total oil in copra. 






Acidity 
as oleic. 


of 
copra. 


Before 
mold ac- 
tion. 


After 
mold ac- 
tion. 


Lobb of oil. 




O. 
4.7734 
6. 1445 
6.3562 


(7. 

1.4749 
1.9646 


O. 
0.8392 


g. 
0.6357 


Per cent. 
43.1 


Per cent. 
21.0 





















66 The Philippine Journal of Science im 

The conditions under which the mold acted were the most 
favorable possible. The coconut meat was kept in a saturated 
atmosphere. The meat was grated, and therefore a greater 
surface was exposed to mold action than is the case in ungrated 
meat. Undoubtedly the loss in oil was higher than is to be ex- 
pected under ordinary conditions. The table is of value also 
in showing that this mold under favorable conditions does de- 
stroy a very high percentage of the oil in the meat and that oil 
from such meat has a high percentage of free acid. 

Grated meat was used to overcome certain difficulties in manip- 
ulation that arose in attempting to discover the oil loss due to 
this mold in migrated meat. It was found possible more easily 
to prevent bacterial growth in grated meat. In ungrated meat 
bacteria soon appear on the surface and change it into a slimy 
mass upon which molds do not grow. 

It is impossible to shred or grind ungrated meat after the mold 
has grown upon it so that complete extraction of the oil is 
possible without a change in the oil content, due to losing meat 
and oil on the grater and thus rendering the results worthless 
for comparison. The noting of the total oil content before and 
after mold action is necessary in order to give the true oil loss. 

ASPERGILLUS NIGKR VAN TIEGH. (BLACK MOLD) 

This species of Aspergillus plays a more important part than 
does Rhizopus sp., but a far less important one than does Asper- 
gillus flavus (brown mold) . It is the black mold often seen on 
badly molded copra. Its moisture requirements are lower than 
those of white mold, but slightly higher than those of brown mold, 
and although it often grows with the latter, it appears only 
upon copra that contains at least 12 per cent of water, and it 
makes its most luxuriant growth upon copra that contains 18 to 
20 per cent of water. Properly dried copra should not have 
such a high moisture content as is required for the growth of 
this mold- However, as copra is usually prepared in the Phil- 
ippines, it often contains enough water for this mold to appear 
and to make a considerable growth, and under such conditions it 
undoubtedly causes an appreciable loss in oil. Our experiments 
indicate that this loss may sometimes be as much as 40 per cent 
of the total oil. 

In hanging drops the spores germinate in about six hours, 
and the subsequent growth is rapid. In forty-eight hours the 
mycelium from a single spore may grow out and extend over an 
area from 3 to 4 centimeters in diameter. On a nutrient agar 
medium such as coconut, prune, or string bean the colonies are 



xii. a. t Brill, Parker, and Yates: Copra and Coconut Oil 



67 



circular in outline and the substratum is yellow. This color is 
also often seen when the mold grows upon coconut or copra, 
and it appears to be characteristic of the species. 

Morphology. — Hyphse hyaline, 3 to 8 ^ in diameter. Stalk of 
conidiophore erect, unbranched, 2 to 3 millimeters in height. 
Head 140 to 160 ^ in diameter. Swollen tip of conidiophore 70 
to 80 fi in diameter. Spores in chains radiating in all directions 
from the tip of the conidiophore. Spores first brownish, later 
black, globular, smooth, 4 to 5 n in diameter. 

Oil loss caused by Aspergillus niger (black mold). — Table VIII 
shows the loss in oil due to the action of this mold upon grated 
copra in a saturated atmosphere during a period of about ten 
days. The experiment was performed and the results calculated 
after the method already described under Rhizopus. 
Table VIII. — Oil loss due to the action of Aspergillus niger (black mold). 



Sample No. 


Weight 
of copra. 


Weight of oil in 
copra. 






Acidity 
as oleic. 


Before 
mold 
action. 


After 
mold 
action. 


Loss of oil. » 




a. 

5.1663 
5.5930 


0. 
1. 8025 
1. 5962 
1.7282 


0. 
0.8778 
1.2282 
1. 6019 


a- 
0.9247 
0.3680 
0.2263 


Per cent. 
51.3 
23.0 
13.0 


Per cent. 
9.5 
8.0 
,9 


2 


S. 





• The wide variation in per cent of oil loss is due to the extent of the mold growth ; for 
example, in the case of sample 1 the meat was covered with a most luxuriant growth, while 
in sample 3 the mold developed a much smaller growth. 



The oil loss varies somewhat in the three samples considered, 
but all show that there is a marked loss of oil under conditions 
favorable for mold growth. The amount of acid in the oil pro- 
duced is too high for the oil to be considered good, but it is 
rather low in comparison with that produced by the brown and 
the white molds. 

ASPERGILLUS FLAVUS LINK (BROWN MOLD) 

This species of Aspergillus is the mold that plays the most 
important part in the destruction of the oil in copra. It is 
the brown mold that is usually seen on badly molded copra. In 
many cases it is mixed with the black Aspergillus discussed above 
and often with the green Penicillium, which is considered later. 
It occurs upon copra with a moisture content of from 7 to 8 
per cent — a water content lower than the average for Philippine 
copra — and because of its ability to grow on copra with so low a 
moisture content, it destroys a very high percentage of the copra 



68 



The Philippine Journal of Science 



of the Islands. The oil destroyed may be almost 40 per cent of 
the total oil contained in the copra. The oil expressed from 
copra upon which this mold has been growing also contains a 
high percentage of free fatty acid. 

In hanging drops the spores germinate in from four to five 
hours. In nutrient agar media the mycelium becomes visible 
in from twelve to fifteen hours and mature spores are produced 
in about forty-eight hours. The growth from a single spore 
is slow as compared with one of white mold or black mold, but 
the early production of spores leads to a rapid multiplication of 
colonies, and so the mold in a comparatively short time will 
completely cover the surface of the meat of half a coconut. 

Oil loss caused by Aspergillus flavus (brown mold). — Table 
IX shows the loss of oil due to brown mold growing upon pieces 
of copra for one month. The experiment was performed and 
the oil loss calculated as follows: Pieces of copra with about 8 
per cent moisture content were weighed, placed in a closed 
container so the amount of moisture in the air could be kept under 
control, and were inoculated with spores of brown mold. The 
total oil content of the copra was determined by the analysis 
of other pieces taken from the same nut. After the mold had 
grown for thirty days, the copra was removed and analyzed. The 
difference between the two analyses gives the loss in oil due to 
the action of the brown mold. At the completion of the exper- 
iment the pieces upon which the mold had grown were covered 
with a dense layer of brown mold from 4 to 7 millimeters thick. 
When broken open, the meat for about one half to two thirds the 
thickness of the piece was light in color, dry and fibrous, and 
in every way the same in appearance as commercial copra dam- 
aged by this fungus. 
Table IX. — Oil loss due to the action of Aspergillus flavus (brown mold). 



l 

2 

4. 
5 





Total oil. 








of copra. 


Original. 


After one 
month. 


0. 


9- 


<7. 


15.57 


9. 1645 


6. 3465 


13.51 


7.9540 


5.5846 


14.83 


8.8386 


6.8570 


18.13 


10. 8055 


6. 7880 


14.89 


' 8.8744 


5.6010 


13.71 


8. 1969 


5. 1610 


11.69 


. 6.7334 


5.3245 


11.66 


6. 7161 


4. 0076 



2. 9816 
4.0175 
3. 2734 

1.4089 
2.7091 



Per rent. 
30.7 
29.8 
33.7 
37.2 
36.8 
37.0 



0.56 

0.56 



Per cent 
6.8 
5.1 
7.7 
7.2 
6.5 
7.0 



Removed before the end of thirty days. 



xii. a.i Brill, Parker, and Yates: Copra and Coconut Oil 69 

The destruction of oil caused by this mold acting for a period 
of one month under favorable conditions is found to be from 
about 30 to 40 per cent of the total oil contained in anhydrous 
copra, while there is a very considerable production of free fatty 
acid, which shows the deterioration in quality of the oil remain- 
ing. As this mold may occur on copra with a fairly small water 
content, a large part of the Philippine copra is subject to its 
attack. Under favorable conditions it may destroy as much 
as 40 per cent of the oil. Since the probable loss caused by 
brown mold averages nearly 25 per cent, it is evident that the 
loss to the Philippines from this cause alone equals a very con- 
siderable portion of the value of the copra exported. 

PENICILLIUM GLAUCUM LINK (GREEN MOLD) 

This is the common green mold often seen upon copra. It 
grows well on copra containing a very low percentage of water. 
Analyses show that this mold destroys hardly any of the oil in 
the copra, and the production of free acid is low ; therefore a good 
grade of oil can be made from copra upon which this mold has 
growth. Its growth appears to be almost entirely superficial. 
It can be readily removed by brushing, leaving a firm white copra. 
A growth of Penicillium alone may be said almost to serve as 
an indicator of good copra, because it grows at a moisture con- 
tent between 5 and 7 per cent, causes practically no loss in the 
oil content, and produces very little free acid. However, green 
mold growing with one or both of the species of Aspergillus loses 
its significance as an indicator of good copra, as it will grow at 
any degree of moisture higher than 5 or 6 per cent, and the 
Aspergillus indicates a high moisture content. The reason green 
mold does not usually appear upon copra with a high degree of 
moisture is because of its slow rate of growth. The spores ger- 
minate in hanging drops in about six and one-half hours, but the 
subsequent growth is slow, and colonies upon copra do not be- 
come visible to the naked eye until from about twenty-four to 
thirty-six hours after the spores are placed upon it. The colony 
grows very slowly, and mature colonies from a single spore are 
hardly ever more than 1 centimeter in diameter. The spores are 
mature in about eleven hours after the mycelium becomes visible. 
Due to the much more rapid growth of the other molds, Peni- 
cillium is either crowded out or covered over by them, and 
under high moisture conditions it is only after the other molds 
have stopped growing that Penicillium becomes visible. 

Oil loss caused by Penicillium (green mold) upon copra. — 
Table X shows the effect of green mold upon the oil content of 



70 



The Philippine Journal of Science 



copra and its production of free acid. Copra dried until it con- 
tained about 7 per cent water was inoculated with green mold and 
placed in a closed container in a moist atmosphere with which 
it was in equilibrium. After the mold had been allowed to grow 
for a period of thirty days, the copra was removed and analyzed. 

Table X. — Oil loss due to the action of Penicillium glaucum (green mold) . 





1 
Total oil. 


Weight 
of copra. 






Before 
mold 
action. 


After 
mold 
action. 


, 


g. 


a. 


13.07 


7. 3192 


7. 3040 


14. 16 


7. 9296 


7.8823 


10.04 


5. 6224 


5. 5731 


13.34 


7. 4704 


7.4306 



After 
Original. mold 
action. 



P.ct. 

0.20 
0.69 
0.87 
0.54 



P.ct. 

0.79 
0.79 
0.79 
0.79 



P.et. 

1.2 
1.0 
0.9 
0.76 



Table X shows that green mold growing under the most favor- 
able conditions during one month causes an almost negligible loss 
of oil and that the production of free acid is extremely low. 

At the end of the experiment the pieces of copra were covered 
with a dense mass of green mold about 5.5 millimeters in thick- 
ness, but when this was removed by brushing, the copra beneath 
was firm and white with no evidence of penetration by the mold. 

SUMMARY OP OIL LOSS AND DETERIORATION CAUSED BY MOLDS 

Table XI, which summarizes the results secured in the ex- 
periments upon oil loss and production of free acid, is included 
in order more easily to compare the loss in quantity and injury 
to the quality of the oil caused by the four principal molds oc- 
curring upon copra. 

Table XI shows that the copra dealer may expect a loss of from 
30 to 40 per cent upon all copra which contains sufficient water 
to enable brown mold to grow; a further increase of the water 
content makes very little difference, as brown mold will then 
grow with black mold and raise the oil loss to that caused by 
brown mold alone. The loss in quantity and quality of oil is 
always in addition to the loss suffered by the purchase of water 
at the price of copra. Where green mold alone is present, there 
is little loss in oil. However, the presence of this mold indicates 
a copra with a higher water content than where no mold is 
present ; if such copra is stored where the air is in a practically 
saturated condition, it will absorb enough water for brown mold 
to attack it much sooner than will clean copra. 



xii, a, 2 Brill, Parker, and Yates: Copra and Coconut Oil 



71 



Table XI. — Suftimary showing effect of mold action on the quantity and 
free acidity of oil in copra. 

[The c intent fur the growth of each mold is g ; ven under the special 

headings.] 





Mold acting for 10 days upon Mold acting for 30 days upon un- 
shredded meat. shredded copra. 


Sample No. 


\<-u-.~ u ! Black mold 

iSSSSSh \ "EST" 


Brown mold Green mold 

(Aspergillus (PcnecUlium 

Jlai'us). glaucum). 




Free 

Oil loss, acid as 

oleic. 


Oil loss. 


Free 
acid as 
oleic. 


Oil loss. 


Free , ' Free 
acid as Oil loss, acid as 
oleic. | oleic. 




Percent. Percent. 


Per cent. 


Per cent. Per cent. Per cent. Per centj Per cent. 
9.5 30.7 6.8 | 6.20 1 1.2 




S6.4 1 19.8 , 23.0 
41.0 26.6 13.0 


3 


2.9 


83.7 7.7 j 0.87 


0.9 

0.8 






1 




36.8 
37.0 
40.3 












7.0 I 
















Average 












40.2 


22.5 1 29.1 
1 


6. 8 1 35. 1 | 7. 1 | 0. 6 


1.0, 



In the experiments to determine the critical .moisture content 
of copra for the growth of the various microorganisms, we found 
uneven distribution of water within a given piece of copra. The 
total water content may be as low as 5 per cent, sufficiently dry 
to prohibit mold growth, while the upper portion of the meat 
may be practically saturated. For clearness and convenience 
the terms upper and lower layer will be used in the following 
discussion. By the upper layer is meant that portion of the meat 
beginning at the surface of the meat adjacent to the water portion 
of the coconut and extending outward approximately to one fifth 
of the thickness of the meat ; the lower layer is the remainder of 
the meat. The moisture content of the upper layer is more im- 
portant in relationship to mold growth than the total water con- 
tent of the sample, because this portion is first attacked. It is 
impossible to obtain exact data as to the water content necessary 
for mold growth, because of the difficulty of obtaining a uniform 
sample of the upper layer. In Table XII the figures are ap- 
proximate values for the moisture content of the separate layers. 

In every case the percentage of water in the upper layer is 
much higher than in the lower layer or in the general sample 
and is high enough to support the growth of the more destructive 
molds. The percentage of water based on the general sample 
would lead one to believe this to be a well-dried copra that would 
be resistant to mold attack. 



72 



The Philippine Journal of Science 



Table XII. — Moisture content of the upper and the lower layers of general 
samples of copra after storage in a moist atmosphere. 



Weight of— 




Water in- 


" 


Upper 
layer. 


Lower 
layer. 


Upper 
layer. 


Lower 
layer. 


General 
sample. 


a. 


g. 


Per cent. 


Per cent. 


Per cent. 


14.7681 


65. 7966 


13.36 


6.4 


7.6 


3. 5100 


43. 8112 


21.52 


7.8 


8.8 


5. 5798 


30. 2842 


20.43 


8.9 


10.7 


14. 1452 


32. 1548 


11.3 


6.5 


8.0 


12. 8122 


26. 5142 


24.0 


6.8 


12.5 


10. 2004 


23. 7875 


10.9 


5.7 


7.9 


11. 0936 


29. 2013 


10.5 


7.6 


8.4 



HEAT AND ITS RELATION TO MOLD GROWTH 

While investigating the rise of temperature occurring in moldy 
copra when piled in large quantities, it was noticed that the tem- 
perature increases to a certain maximum of about 50° C. and 
then declines. As this rise in temperature appeared to be re- 
lated to mold action, the question presented itself whether the 
fall after attaining a maximum might to be due to the fact that 
such temperatures inhibit the growth or kill the mold concerned. 
It is well known that temperatures around 50° C. will, when 
long maintained, kill the mycelium of many species of fungi or 
at least stop their growth. In general, the spores of fungi are 
more resistant to heat than the mycelium and so might survive 
such temperatures. However, in this case there was no recur- 
rence of a high temperature, as might have been expected had 
spores of the fungi survived and, after the temperature decline, 
returned to normal germination and resumed their growth. To 
determine the effect of such temperature upon the spores and 
mycelium of the four molds considered in this paper, the follow- 
ing experiments were conducted: 

Molds in pure culture were allowed to grow until a mycelium 
was distinctly visible and then were placed in an oven maintained 
at a constant temperature of 50° C. At intervals cultures were 
removed from the oven and kept at room temperature to see 
whether growth was resumed. To determine the effect of this 
temperature upon mold spores, pieces of coconut meat were in- 
oculated with mold spores, put in closed containers, and placed 
in the oven. At intervals pieces of the meat were removed and 
kept at room temperatures to determine whether germination 
occurred. The observations from these two series of experi- 
ments indicate that both the spores and the mycelium of each 



xii. a. s Brill, Parker, and Yates: Copra and Coconut Oil 73 

of the molds survive a temperature of 50° C. for from nine to 
twelve hours, since removal from the oven permitted both ger- 
mination and growth to proceed at the normal rate. 

ABSORPTION OF WATER BY COPRA 

The question may be asked, will copra once thoroughly dried 
absorb enough moisture from the atmosphere during storage to 
raise its moisture content to a point where molds may grow? 
Should this occur, what is the use of thorough drying, if during 
subsequent storage and shipment mold will gain a foothold and 
destroy a high percentage of the oil and impair the quality of 
the remainder? This is a most important question from the 
point of view of the copra dealer, and in .order to formulate an 
answer, a number of variable factors must be taken into con- 
sideration. 

Generally speaking, it has been our experience that when 
dried to a moisture content of about 6 per cent copra does not 
mold when stored where there is a circulation of air. However, 
when kept in a saturated atmosphere, it will eventually absorb 
enough water to enable mold to grow, but this only takes place 
after it has been exposed to such conditions for from one 
to two weeks. It would probably very rarely happen that the 
air in a warehouse or in the hold of a ship will reach a point 
of saturation and, if this does occur, that the humidity will be 
replenished when it is absorbed by the copra and that these con- 
ditions will last for so long a period even as one week. Copra 
which has been dried only to about 7.5 per cent of moisture often 
will not grow mold in the open air probably because the surface 
is really dried out below this moisture content, but when stored 
under the same conditions as above, will develop mold in from 
two to five days. 

Another factor which has an influence is the role played by 
the water liberated during the processes of metabolism of the 
mold. Mold grows at the expense of the oil and fiber of the 
copra, converting them into water and carbon dioxide. If the 
lower part of a large pile of copra contains sufficient water to 
permit molds to grow, the whole may mold, due to the water 
that is liberated during mold growth being confined in the 
pile long enough for the remainder of the copra to absorb enough 
moisture in turn to support mold growth. 

All of our experiments tend to show that copra once properly 
dried to approximately 6 per cent moisture does not absorb suf- 
ficient water, unless in a saturated atmosphere for prolonged 
periods of time, to develop even a superficial growth of mold. 



74 



The Philippine Journal of Science 



In Table XIII weighed pieces of copra were kept in a saturated 
atmosphere at room temperature with daily weights recorded 
until mold growth appeared. 

Table XIII. — Change in moisture content of copra stored in saturated 
atmosphere {at 29° C.) until mold appeared. 



Original weight - grams. 

Total increase in weight per cent. 

Original total water -do 

End total water - do 



Gain in weight of sample 



8. M 

3.64 
4.24 



0. 24 
0.71 



2.25 
2.30 
2.87 
2.73 



In one case mold appeared on the eighth day, in two on the 
tenth, and in the remaining one on the twelfth. 

Table XIV. — Change in weight of copra stored in open air at atmosphert 
temperature. 



Relative 
humidity 

in Manila 
on day 

weighed.* 



94.5 
73.5 
94.2 
88.6 



Total lose.. per cent.. 

Original total water... do 

End total water do 

Original copra weight grams.. 

End upper portion water per cent.. 

End lower portion water do 




• Algue, Jose. Weather Bureau Bulletins for June and July (1916). 

» This is marked +, where an increase in weight resulted, but no mark is used where this 
change was a decrease. 



xii. a, i Brill, Parker, and Yates: Copra and Coconut Oil 75 

In Table XIV are recorded data relative to weighed pieces of 
copra kept in the open. The weight in grams is recorded under 
series number of samples. 

At the end of fifty-six days no mold growth had formed. The 
last ten days were a period of wet weather, during which mois- 
ture was absorbed by the upper portion of the meat as shown 
by comparison of the upper and lower portions. This change 
of moisture content of the upper layer when calculated on the 
total weight shows a relatively small change in the percentage 
moisture. From this it is seen that copra stored in the open air 
will lose or gain water until it is in equilibrium with the at- 
mospheric moisture. This equilibrium point is about 5 per cent 
moisture for copra. 

The copra in Table XIII had developed a fair growth of mold 
by the end of two weeks. There are only short periods during 
the rainy season when the air is practically saturated with mois- 
ture, and unless dry copra were stored in containers in a 
saturated atmosphere, no molding would take place. It will be 
noticed that the total moisture content of each piece is extremely 
low, but if, on the other hand, copra containing from 7 per cent 
to the amount of moisture contained in fresh meat had been 
stored under like conditions, the mold would have practically de- 
stroyed the copra in the course of several days. This would also 
be the case if copra of 7 per cent or more was stored without 
air circulation, the damage due to mold depending upon the 
moisture present. In Table XV data on good quality of machine- 
dried copra without mold, stored in five sacks, are recorded. 
Temperature readings of the copra and outside atmosphere were 
noted from time to time. Weights of the lot before and after 
were not recorded, as insects destroyed a part of the copra. 

Table XV indicates no temperature change where no deteriora- 
tion of meat and oil occurs and, further, that copra once properly 
dried does not develop mold when stored. 

Copra which has been machine-dried and not cooled before be- 
ing placed in a large pile becomes hot and later often shows mold 
growth. We believe this to arise from a breaking down of oil 
and cellular matter from heating (the heating is similar in 
character to the spontaneous heating of oily rags) into carbon 
dioxide and water vapor. The moisture of the surface of the 
copra is raised in this manner to a point where it will support 
mold growth, and thus microorganisms appear. The temper- 
ature at which the copra is stored is a factor in this heating. 
When the copra is carefully cooled, the oxidation at the lower 
temperature is so slow that no appreciable heat is given off and 



76 



The Philippine Journal of Science 



it escapes without any rise in the temperature of the pile. The 
theory of some of the copra men that the moisture content of 
the copra is increased by the condensation of moisture from 
the cool air surrounding the warm copra is absolutely untenable. 

Table XV. — Data on machine-dried copra stored in sacks. 



Date. 


I 

Copra. 






Sample. 




Air. 

°C. 

31 

30.5 

30.5 

30.5 

31.5 

31.5 


II. 


III. 


IV. 


V. 


Copra. 


Air. 


Copra. 


Air. 

°C. 

29.5 

29.0 

30.0 

31.0 

31 

30.5 


Copra. 

°C. 

30 

30.5 

29.5 

32.0 

29.1 

31.5 


°C. 

30 

30 

30 

32.5 

28.5 

31.5 


Copra. 


Air. 




°C. 
31 
30 
30 
31 
32 

•6.1 
4.9 


°C. 
30 
29 
28 
29 
29 
29 


°C. 

30 

30 

29.5 

29 

29.5 

29.5 


°C. 
29.5 
29.0 
30.5 
31.0 
30.5 
31 


°C. 

30.5 

31.0 

29.5 

32 

32 

32.4 


°C. 
30.5 
31.0 
30.0 
32.0 
81.0 
31. S 












Original water P. ct.. 

Water May 15.... do.... 




5.9 




4.8 




4.1 




4.1 





Moisture content of a general sample made up of I. II, III, IV, and V. 



COPRA-DRYING METHODS IN USE 

The poor quality of Philippine copra due to insufficient drying 
and improper handling is, of course, dependent upon the methods 
employed. There are two general methods in use in the Islands, 
sundrying and kiln drying; improvement of both is possible to 
a certain extent. The sundrying method used throughout the 
southern islands produces the better grade of copra ; it consists 
simply in halving the nuts, without previously husking, and 
exposing the meat to the sun. Where sufficient care is exercised 
in the way of cleanliness and complete drying, the method pro- 
duces an excellent grade of copra. 

The periods of daily rains do not favor this process in parts 
of the Philippine Islands, and there the grill method is used. 
In the grill, or tapahan, process the nuts are husked, halved, and 
placed on bamboo mats, under which shells and husks are 
burned. 15 After the meat is partially dry, it is removed from the 
the shells and is subjected to further drying. Analyses show 
that the finished product as it leaves the dryer contains at least 
20 per cent moisture. 

" For more complete description of sundrying and grill methods, §ee 
Pratt, D. S., Thia Journal, Sec. A (1914), 9, 177. 



xii. a. 2 Brill, Parker, and Yates: Copra and Coconut Oil 77 

Table XVI. — Water content of tapahan-dried copra. 

1 21.57 

2 25.57 

3 28.77 

4 25.98 

5 20.52 

6 25.54 

7 28.51 

The samples noted in Table XVI were obtained from the 
copra from seven tapahans at the completion of the drying 
process. The very high moisture percentage indicates imper- 
fect drying, a condition most favorable for mold growth and 
bacterial action with the attending deterioration of the copra. 
After the first handling and drying to only approximately one 
half the original moisture contents, these samples were dirty 
and smoky in appearance. Complete drying by the tapahan 
method, a proceeding probably never carried out in practice 
by the producers, requires about four days and gives a very 
•badly smoked product. The uneven heating obtained on the 
tapahan is a bad feature of the method in that the pieces directly 
in contact with the grill are overheated and dry much more 
rapidly than the upper layers. This is especially true in cases 
where the half nuts are piled up several layers deep. The ta- 
pahan is so constructed that it favors uneven drying and much 
smoking of the copra. It consists of a pit, dug in the ground 
to the depth of 3 meters, connected at the bottom by a narrow 
underground tunnel to the heat chamber. The heat chamber, 
is approximately 2 meters wide and 6 meters long, the dimen- 
sions varying, with the bottom sloping down to meet the tunnel. 
It is covered with a plaited mat of bamboo or rattan, or in some 
cases with stripes of split bamboo, upon which the half nuts are 
piled. Shells, and sometimes husks as well, are burned in the 
bottom of the pit, the heat and smoke passing through the pile 
of copra. An attempt was made to improve a "tapahan" by 
providing it with a chimney and a special fire box. By burning 
only shells, much of the smoke was eliminated, but the copra 
was unevenly dried. 

Even with these two existing methods a better grade of 
copra could be prepared than is now the case in the Philippines if 
proper care were exercised in handling the product. By using 
sundrying in conjunction with the tapahan, a more evenly dried 
product could be obtained. In the southern islands, where the 
drying could be carried on entirely by the sun's heat, the 
precautions necessary would be to keep the copra free from 



78 The Philippine Journal of Science wx 

dirt and to secure more complete drying with facilities to protect 
it during short rainy periods. Pratt J0 states that copra produced 
in other countries by like means commands a higher price than 
Philippine copra and must be considered as superior in the 
world's market. 

MECHANICAL DRYERS 

Mechanical dryers are one possible solution of the problem 
for improving the copra production of the Islands, but they 
have not been introduced in a commercial way. Several ma- 
chines have originated in the Philippines that require consider- 
able handling of the meat, but while their product is of good 
quality, initial and operating expenses seem to be too great to 
warrant their adoption. Smith 1T gives descriptions and draw- 
ings of several types of drying machines. At two of the copra- 
oil mills here large hot-air dryers have been installed for redry- 
ing copra before it is milled. This apparatus, with changes, 
might be adapted to the drying of fresh meat. It has the ad- 
vantage over other forms of dryers in being a continuous process 
with the minimum amount of handling of the product. A small 
cheaply constructed dryer, combining low operating expenses 
and rapid drying, would find a ready market in the Philippines. 

Mechanical drying machines have been discredited by the pub- 
lic through the belief that when coconut meat is dried in a 
current of hot air a part of the oil is carried away from the 
copra. This supposition is not in accord with what one might 
expect from a study of the physical properties of coconut oil. 
It was found that when coconut oil was heated for four hours 
in a current of hot air at 100° C. there was no appreciable loss 
in weight. 

Pure coconut oil was dried in vacuum over sulphuric acid for 
four days, after which it was heated in an oven at 100° C. for 
four hours with the following results : 

Table XVII. — Loss on heating coconut oil at 100" C. for four hours. 



Oil No. 


Weight of 
oil. 


Weight 
lost. 




0. 

i. 9112 
4.S919 
3.9282 
4. 7955 


Per cent. 

0.024 
0.022 
0.023 






1 IV 


1 





11 Op. cit. 

" Smith, H. Hammel, Coconuts. The Consols of the East. Tropical Life, 
London (1913). 



xii, a,* Brill, Parker, and Yates: Copra and Coconut Oil 



79 



However, it did appear possible that there might be a mechan- 
ical carrying over of the oil with the escaping moisture, espe- 
cially if the drying temperature were high and the driving off 
of moisture rapid. Experiments were made by examination 
of the gases from a mechanical dryer. The type of machine 
used was similar to the ordinary oven dryer, but provided with an 
arrangement for the circulation of hot air. The experiment 
was carried out in the following manner : 

A weighed amount of coconut meat was distributed over the 
trays and dried at various temperatures. The volume of air 
passing through was calculated, and the temperatures at the 
entrance and exhaust were noted. An aliquot part of the air 
was withdrawn from the oven by means of a vacuum pump at 
the rate of 1,200 liters per hour, measured by means of a gas 
meter. " The air in passing to the pump and gas meter was con- 
ducted through two series of four bottles, each containing chloro- 
form immersed in a freezing solution in order to free the air of 
oil. Four separate determinations were made, using drying 
temperatures varying from 70° to 100° C, and in no case was 
n qualitative test for oil obtained. 

Tablb XVIII. — Data on oil loss in mechanical dryer. 



Tem- 
per- 
ature. 


Freih 
meat. 


Weight 
loss. 


°C. 


Kilo*. 


P. cent. 


70 


100 


40.0 


80 


104 


42.0 


90 


102 


41.0 


100 


86 


44.0 



Determinatio* 1 ...'. 

Determination I . 

Determination J 

Determination 4 I 



Per cent. 
0.0 
0.0 
0.0 
0.0 



The results recorded in Table XVIII are opposite to the belief 
of the Ceylon planters, as reported by Pratt : 1S 

The amount of copra from a given quantity of fresh nuts depends to a 
considerable extent upon the rate of artificial drying. Ordinarily, from 170 
to 200 nuts give about 50 kilograms (110 pounds) of copra. The two 
extremes are encountered in comparing the output of sundried copra with 
that of desiccated coconut products. * * * The decrease in time re- 
quired for expelling the water is, therefore, coincident with increased Joss 
of oil, and all methods of preparing copra must represent an economical 
balance between these factors. It is unquestionably possible to make copra 
in much less time than is required by either the sundrying or grill-drying 
processes, but experiments made by planters in Ceylon have not impressed 
them with the advisability of adopting such changes. One of the most 
progressive coconut planters in the island constructed a drying house with 
brick heating flues and produced paper-white copra in less than twenty-four 



Pratt, D. S., This Journal, Sec. A (1914), 9, 181. 



80 



The Philippine Journal of Science 



hours, but discontinued the process because of the resulting high loss of 
oil. It is his opinion that a continuous slow current of air at from about 
54° to 60° (130° to 140° F.) — the proper temperature to be determined by 
experiment — should complete the drying process within three days and 
nights, and with the least loss of oil. A rapid drying in ten hours must be 
attended by a considerable loss, and will require about 15 per cent more 
kernel to produce a given weight of copra. 

With several of the statements in the above-quoted opinion 
we must take issue. Our experience proves that no oil loss 
occurs when rapid drying takes place at 80°, 90°, or 100° C. If 
by the ambiguous statement, "a considerable loss, and will re- 
quire about 15 per cent more kernel to produce a given weight 
of copra," is meant anything other than more complete drying, 
the experience of the Ceylon planters is totally different from the 
results obtained in this investigation. As further evidence of 
the high yields of oil from desiccated coconut, Table XIX is 
appended. Did such losses of oil as reported by Pratt occur, 
the yields of oil in the copra recorded in this table would be 
much lower. 

Table XIX. — Data on drying copra under various conditions. 



Sample 
No. 


Means of drying. 


Oil in an- 
hydrous 
material. 


Water. 


1 
2 

3 
4 
5 

6 

7 
8 
9 
10 
11 
12 
13 
15 
16 




Per cent. 


Per temt. 


do - - - 


74.10 
70.7 
71.3 
73.7 
69.1 
72.9 
68.0 
71.5 
71.2 
70.8 
69.9 
72.2 
68.5 
69.0 


6.4 
7.3 
7.1 
4.0 
5.2 
3.9 
4.7 
5.3 
.9 
6.8 
9.1 
5.4 
1.1 
7.5 






do - - --. 




do 


do 


do 















PREPARING COPRA BY USE OF SULPHUR DIOXIDE 

The Bureau of Science has developed a simple method for the 
preparation of copra by treatment with sulphur dioxide gas and 



" Since perfecting this method we have noted the sulphur dioxide method 
of Marot (English patent 6379, 1906) described by Lewkowitsch in Chemical 
Technology and Analysis of Oils, Fats and Waxes. 5th ed. MacMillan & 
Co., London (1914), 2, 631. 



in. a. 2 Brill, Parker, and Yates: Copra and Coconut Oil 



81 



allowing the meat to dry without the addition of artificial heat. 
The apparatus used is a wooden box, provided with trays with 
split bamboo bottoms, and a 4-wheel car consisting of framework 
and two pairs of small iron wheels mounted on a wooden track. 
The box is 1.05 by 2.10 by 2.70 meters and is provided with a 
door in one end, being of sufficient size to accommodate twelve 
trays separated sufficiently for free circulation of the gas and 
with a capacity for holding about 1,200 nuts. The track is twice 
the length of the box to facilitate loading and unloading of the 
car outside of the box. The operation is carried out in the 
following manner : The husked nuts are halved and spread on 
the trays with the concave side down, and the trays are loaded 
on the car and pushed into the box. Eight kilograms of sulphur 
are next placed in a shallow pit under the car and ignited. If 
the box is made comparatively tight, this amount of sulphur will 
burn for from ten to twelve hours, liberating sufficient sulphur 
dioxide gas for the treatment. At the end of the sulphuring 
period the car is rolled into the open and the meat is removed 
immediately from the shells with an ordinary copra knife, or is 
allowed to remain in the shells for from four to five days, at 
the end of which time the meat has become sufficiently dry to 
allow of its ready removal. In either case the sulphured ma- 
terial is spread out under cover with free access of air for a 
period of two weeks, after which it is cut up and sacked for the 
market. Obviously on sunny days spreading the meat in the open 
greatly hastens the drying and should be taken advantage of. 

Table XX. — Sulphur dioxide and water content of copra at various periods 
during the sulphur treatment.'' 



Sulphur dioxide. 



Hours Content Loss in 

*'-*" dXlvi Content total 
lft!l after 10: content 

treat days ' ' dur ' ne 

rnlnt. 10 da ^" 



cd. 



^Percent. Percent. Percent. 



6 0. 1218 

9 0. 1456 

13 i 0.1869 

18 , 0.2114 



0. 0412 



24 



.3144 



0. 0425 I 
0.0672 
0.0722 I 
0.0715 I 



66. 2 
70.8 
64.1 
65.9 



Loss 



Imme- 
diately 
after 
treat- 
ment. ! ment. 



! !' 

Loss in ; 
i„w!!l Ten days total. 10 
'?*£*'■ after days 
treat ' treat " ' after 



ment. 



Percent. Percent. Percent. Percent. Percent. 



47.7 
46.1 
49.2 



37.2 
39.0 
45.5 



14.2 
15.5 
27.0 
23. 
15.9 I 



13.1 
12.6 
13.7 
11.1 



73.1 

72.9- 
73.0 



11 The maximum amount of sulphur dioxide that coconut meat will absorb was not determined. 
However, a twelve-hour continuous treatment with the gas was found to be- sufficient in 
all cases. 



82 The Philippine Journal of Science iw 

The action of the sulphur dioxide on coconut meat serves a 
double purpose. The mold spores are destroyed, and a sufficient 
amount of the sulphur dioxide remains in the meat for the time 
necessary for the escape of the water to prevent the growth of 
new mold spores. Also, as will be shown in Table XX, there 
is a considerable loss of water during the treatment due to the 
softening of the cell walls from the action of the sulphur dioxide, 
which has a marked affinity for water. 

Table XX shows data taken from analysis of sulphured coconut 
meat which was allowed to dry under cover. At no time was 
this meat exposed to the drying influence of the sun or artificial 
heat. The figures are averages from ten samples for each period 
of treatment and represent a close approximation of the true 
sulphur dioxide content. 

The residual sulphur dioxide after one month's time may be 
disregarded, as only traces are found. A part of the samples 
from the series receiving six and nine hours sulphuring developed 
a slight growth of green mold which was entirely superficial, 
and after the meat was completely dry, it could be brushed off, 
leaving white copra exposed. Copra prepared by this method 
with treatments of twelve hours or more and if dried within 
two weeks is clean and white and free from mold, the oil ex- 
pressed being practically colorless and free from rancidity and 
acidity. Sulphur dioxide was not detected in the expressed oil. 
It is pronounced equal to, or better than, the best Cochin oil and 
will, therefore, sell for from 4 to 6 centavos a pound more than 
the ordinary oil. At 4 centavos per pound this would mean an 
increased value of 3.28 pesos per picul of copra. With proper 
organization the labor cost by this method should not exceed 
that of the tapahan or sun-drying processes, and the initial out- 
lay is small. Tarred paper may be substituted for wood in the 
construction of the box, and bamboo for the trays is inexpensive. 
Unless sundrying is used in conjunction, two weeks should elapse 
before marketing the copra, and while this time element is an 
objectionable factor, the increased value of the product would 
more than pay for the delay. The expense of sulphuring is also 
negligible when the superior quality of copra obtainable is con- 
sidered. The sulphur need not be a refined grade; in fact the 
crude sulphur is preferable, as the impurities slacken the rate 
of combustion and, hence, increase the length of time of sulphur- 
ing from a given amount of sulphur. Sulphur is available in 
both the Philippines and Japan. A local demand would develop 
the sulphur resources of the Philippines; otherwise it could be 
imported from Japan more cheaply than is now the case with 



in. a, 2 Brill, Parker, and Yates: Copra and Coconut Oil 



83 



Japan sulphur in the United States. In 1915 the United States 
Geological Survey :o reports that during the calendar years 
1911-14 Japan exported to the United States 16,185, 24,505, 
15,317, and 21,913 tons of sulphur, respectively. If all of the 
nuts grown in the Philippines, approximately 431,387,000, were 
made into copra by the sulphur dioxide process, not more than 
3,500 tons of sulphur would be required annually. 

Pure coconut oil is water white and free from acidity and 
rancidity. It cannot be obtained from copra which has been 
acted upon by microorganisms. We would define good copra as 
coconut meat containing from 4 to 6 per cent water and 65 per 
cent oil, calculated on anhydrous meat, with an acid content of 
less than 1 per cent acid (oleic) and free from dirt and smoke. 
Such copra would find top prices on the world's market, and 
there is no reason why with time and improved methods Philip- 
pine copra could not be the world's standard rather than the 
poorest quality. The increased money value of such a product 
can be readily seen from a consideration of weight losses in 
storage and during shipment. Further the oil obtainable due 
to lack of necessity of purifying it from free fatty acids, ran- 
cidity, and color add materially to its value. The cost of refining 
commercial coconut oil is estimated at 8 pesos per each 1 per 
cent fatty acids per ton of oil. 

Below are given comparative figures of market quotations for 
best-quality raw co:onut oil and the ordinary commercial product. 

Table XXI. — Comparison of prices in pounds sterling per ton of various 
grades of coconut oil in the London market.' 1 



Date. 


Oil. 


Cochin. 


Ceylon. 


Commer- 
cial copra. 




52-54 

60 

51-53 


49-51 

50 

49-51 


45-51 
46-49 
38-48 









'Tropical Life (Dec, 1915), 236. 

FURTHER SUGGESTION FOR THE IMPROVEMENT OF COPRA 

Ripe nuts only should be used for the production of copra, be- 
cause only fully matured nuts produce copra containing the maxi- 
mum oil content. 21 The practice of picking green nuts in the 



Cox, Alvin J., Bureau of Science Press Bull. (1916), No. 54. 
See Walker, H. S., This Journal (1906), 1, 58. 



84 The Philippine Journal of Science 1917 

Philippines in order to take advantage of market fluctuations is 
well known, and sorting out the "mulos" — rubbery pieces of un- 
ripe meat, which will not become dry — is a recognized part of the 
copra industry. Aside from its low oil content, green copra 
must be worked separately; consequently the presence of 
green copra interferes with the routine of the mill and increases 
the cost of handling. In some' coconut-producing countries laws 
are or have been in effect prohibiting the picking of coconuts. 22 
Such a law is impracticable in the Philippine Islands, since cer- 
tain kinds of coconuts here do not drop their fruit until they 
have become much overripe. In order to make it practicable 
to find and collect nuts which have fallen from the tree, clean 
cultural methods should be practiced on the plantation by burning 
the taller weeds and dead wood. By such methods of culture not 
only is the soil fertilized to a certain extent by the ash and 
decaying vegetation, but also the breeding places of many of 
the insects and fungi which prey upon the coconut are destroyed, 
so that the health of the trees is improved and ultimately the 
production of nuts increased. 

The nuts when opened should be kept free from dirt and dried 
immediately and without smoking to a water content of 5 per 
cent. 

The establishment of copra standards and organization of 
copra centrals by the Government has been proposed. The 
Bureau of Science has for some time had the problem of classifi- 
cation of copra under consideration. In view of the large num- 
ber of classes necessary to meet the requirements due to the 
several factors — oil content, moisture, acidity, and appearance 
of the copra — complete standardization would be difficult to 
establish. 

Under the present conditions a consignment of copra received 
from the provinces at the Manila bodegas may represent the 
product of 100 or more different producers, which means that 
practically no two sacks have been dried to the same moisture 
content. And aside from the necessity of analyzing a sample 
from each sack, the fact that. uneven drying has been practiced 
and that the variation of the several factors in each nut is con- 
siderable, the analysis would be at best a rough approximation. 
The classification of copra under chemical control would be pos- 
sible in conjunction with a system of inspection, as is practiced in 
hemp grading. It requires experience in order only to tell if a 
piece of copra is dry. In the hands of honest and competent 

" Pratt, D. S., op. cit., 179. 



xn.A, 2 Brill, Parker, and Yates: Copra and Coconut Oil 85 

inspectors stationed at the various shipping centers in the prov- 
inces a uniform product only would be marketed, and under 
this condition chemical analysis would be a valuable means of 
determining the worth of a parcel of copra. 

The probable solution of the copra problem will come' through 
educational measures and encouragement of the investment of 
capital in coconut plantations. Modern plantation managers will 
• apply improved methods, and the small grower will necessarily 
follow their example or be satisfied with a discriminating price 
for his poorer quality of copra. At present the Manila market 
does not demand a high grade of copra. Any quality of copra 
sells and with too little difference in price from the best grade. 
The ordinary dealers are satisfied to make their profit on poor 
grades of copra. At present a poor grade of copra controls the 
market; in fact no other is available in large quantities, and 
parcels of good copra are so rare that they must be sold at the 
normal price and thrown in with the bad. Just as soon as 
there is a sufficient amount of good-quality copra offered to 
pay for handling it separately, a new standard will be set and 
sufficient discrimination shown between good and poor copra to 
pay the producer for extra care in producing that of good quality. 

Three grades of copra are regularly recognized in the Manila 
market : Cebu Sun Dried, Fair Marketable Manila, and Laguna. 
A discriminating price of about 75 centavos per picul exists for 
Sun Dried over Laguna. 

The above condition is applicable to the usual local conditions 
and is not true of the world's market, where a discriminating 
price is paid for copra dependent on the quality. Inquiries have 
been received by local merchants asking for terms for the supply 
of high-grade copra in large quantities. The price quoted was 
sufficiently higher to warrant the use of greater care and the 
expenditure of considerable more money in the preparation of 
the copra. However, under present conditions it is impossible 
either to buy large quantities of high-grade copra in the local 
market or to produce the same with existing methods. One 
shipment of 1,000 tons of high-grade copra from the Philippines 
would do much toward raising the present standard of our copra 
in the world's market and to increase the demand for it. 

SUMMARY 

The moisture content of representative samples of commercial 
copra is shown. Tables showing the loss in weight, the rise 
of temperature, and the presence of notable quantities of carbon 
dioxide, in the surrounding atmosphere when poorly dried copra 



86 The Philippine Journal of Science 

is stored, are included. Well-dried copra does not show this 
phenomenon of oxidation when stored. 

A distinction is made between the true and the apparent loss 
in oil when copra molds. 

A determination of the moisture content necessary for the 
growth of the four common molds — the white, black, brown, and 
green — has been made, and the loss in oil resulting from their 
action on copra for definite periods of time has been found. A 
description of their habits of growth is included. Data are 
presented which show that a considerable difference exists in 
the moisture content of the upper and lower layers of the copra 
and that the upper layer may contain sufficient moisture for 
mold growth when the moisture content of the general sample 
is below the moisture requirement for this growth. 

The local copra-drying methods are discussed, and proof is 
offered of the nonloss of oil when copra is dried in a hot air 
drier at temperatures ranging from 70° to 100° C. 

A method making use of sulphur dioxide has been investigated, 
and the results are presented. 

Methods for the improvement of Philippine copra are sug- 
gested, and the difficulties of making changes in the present 
methods of marketing are considered. 



METHODS FOR THE PRODUCTION OF PURE COCONUT OIL ' 
By Harrison 0. Parker and Harvey C. Brill 

(From the Laboratory of Organic Chemistry, Bureau of Science, 
Manila) 

The production of copra has always been the most convenient 
means of handling coconut, since it eliminates the shells and husks 
from transportation charges and the cost of installation and 
operation of oil mills. Little attention, particularly in the Phil- 
ippines, is paid by the producers to the quality of the copra con- 
cerned in its relationship to the oil and press cake obtainable. 
This laboratory has been considering the possibilities of obtain- 
ing pure oil direct from the fresh coconuts, thereby eliminating 
the necessity of turning the nuts into copra with its attendant 
deterioration, loss of oil, and the refinement of a product which 
might be originally prepared in a pure condition. 

A crude method of obtaining oil from fresh coconuts as prac- 
ticed in the provinces is as follows. The nuts are husked by 
breaking on the sharp point of a plowshare, halved with bolos, 
and the meat is removed and grated in one operation by means 
of a steel burr driven with pedal attachment. The same operator 
holds the half nut and drives the burr. The grated meat is 
transferred to the caua, or steaming kettle, where it is mixed with 
one half its volume of water and steamed for from two to four 
hours by the application of direct heat from burning shells and 
husks. The steamed meat is then placed in rattan bags, which 
are suspended in a perpendicular position between two heavy 
pieces of wood, and pressure is applied by means of a wooden 
vice screw. The gata, or white emulsion of oil, water, and 
cellular tissue obtained, is returned to a second caua, where the 
water is evaporated and the cellular tissue coagulated into a 
brown mass. This cellular tissue, of high protein content, is 
used by the Filipinos as a food. The oil is ladeled into earthen- 
ware vessels, and the small solid particles remaining are allowed 
to settle out. No filtering* process is resorted to. The oil so 
obtained is used locally for edible purposes. The press cake is 
shaken through a bamboo basket in order to disintegrate it and 

1 Received for publication February, 1917. 

87 



88 The Philippine Journal of Science 1917 

is then allowed to ferment for a period of three days, with oc- 
casional stirring and turning of the meat to hasten fermenta- 
tion. After the first fermentation period the meat is ground by 
means of the iluhan, a crude apparatus consisting of a heavy 
roller working back and forth over a flat surface, and is then 
repressed. The process of fermenting and repressing is carried 
on daily for from five to seven consecutive days. All of the oil 
obtained from the various pressings of fermented meat is classed 
as rancid oil for soap making and similar articles. The poonac, 
or press cake, is used for feeding hogs. 

The edible oil represents about one third of the total oil 
obtained and is of fair quality, having some fatty acids and a 
slight burnt taste and odor and not being entirely free from 
color. It commands a somewhat higher price than does the 
rancid oil. The latter, besides being rancid, is highly colored and 
contains a high percentage of free fatty acids. 

Below are given data showing the amounts of oil remaining 
in the meat after the various pressings, acid contents of oils, 
and cost and sale prices of the oil at Pagsanjan in September, 
1915. The labor cost for milling the grated coconut meat is not 
included in the calculation nor is any value assigned to the press 
cake, although it still contains about 24 per cent oil. ' These items 
probably nearly balance under the system used by the Filipinos. 

Table I. — Data on the Filipino process for obtaining oil from the fresh 
" , coconut. 
Press cake analysis: 

After 1st expression (unfermented meat) 
After 2d expression (fermented meat) 
After 4th expression (fermented meat) 
After 5th expression (fermented meat) 
After 7th expression (fermented meat) 
Acidity of oils (oleic) : 
Edible oil from evaporated emulsion 
Rancid oil after 5th expression of rancid meat do. 
Cost price of oil from 1,000 nuts: 
Nuts 
Husking 
Grating 

Total 
Selling price of oil from 1,000 nuts: * 
Edible oil 37.85 liters (10 gallons) 
Rancid oil 75.70 liters (20 gallons) 

Total do. 30.50 

Net profit from oil from 1,000 nuts do. 7.50 

* One peso Philippine currency equals 100 centavos, equals 50 cents United States currency. 



Oil in 


press cake. 


per cent 


44.18 


do. 


36.68 


do. 


30.68 


do. 


27.69 


do. 


23.90 


do. 


0.6 


at do. 


9.5 


pesos a 


20.00 


do. 


.50 


do. 


2.00-2.50 


do. 


23.00 


do. 


10.50 


do. 


20.00 



xii. a, ■> Parker and Brill: Pare Coconut Oil 89 

Mr. 0. Vyner, of British North Borneo, reports a recently 
patented method for obtaining oil from fresh coconuts, which 
involves the use of modern machinery and which he claims can 
be extended advantageously to a daily capacity of 10,000 nuts. 
The following is a brief description of the process as outlined in 
his letter to the Bureau of Science. 

The meat is removed from the shell by means of a small, one man 
operated machine, with a capacity of 25 nuts per hour. The meat is then 
carried on an endless chute to roller crushers and then into the grinder where 
it is rendered into a fine state of division. From the grinder it passes 
into a large tank called a receiver, where it is thoroughly mixed with a 
definite quantity of pure water [amount not given]. The mixture is then 
run into the boiler where it is heated to a certain temperature and the 
temperature maintained for a period of time [temperature and time control 
not given]. The boiled material is next allowed to flow over a coarse 
strainer, the emulsion passing into a container and the pulp transferred to 
the molder, where the meat- is made into cakes for the final pressing. . The 
liquid from the molder also flows into the container. The contents of the 
container are pumped into the evaporator, where they are heated [tem- 
perature not given] until the oil separates from the rest of the liquid, when 
it is drawn off as the finished product ready for storage. The pulp from 
the molder is next subjected to high pressure resulting in the elimination 
of the water from the "poonac" so formed. 

We were unable to duplicate these results in laboratory prac- 
tice, probably due to the fact that a complete description of the 
process was not given. 

We found by repeated trials that freshly grated coconut meat 
would not give up its oil readily by pressing. When subjected 
to a pressure of 70 kilograms per square centimeter (1,000 pounds 
per square inch), over 60 per cent of the total oil remained in 
the press cake. By treating the meat with water and live steam 
for a period of three hours before pressing, 80 per cent of the 
total oil was removed by one pressing. A mechanical agitator 
used in the above experiment, which beats up and thoroughly stirs 
the meat and water during the heating, brings more of the oil in 
contact with the liquid through breaking of the cell walls, so 
that a greater amount of oil is obtained when the mixture is 
pressed. The liquid obtained after pressing is a white emulsion, 
consisting of cellular tissue, oil, and water. Oil prepared from 
the emulsion by heating over a direct flame as described in the 
native process is colored and possesses a burnt taste and odor 
even when small amounts of the emulsion are evaporated. The 
quality of the oil was not improved by evaporating the water 
from the emulsion under diminished pressure, since the jellylike 
mass of cellular tissue and oil remaining must be further heated 
at a higher temperature completely to coagulate the tissue in 



90 The Philippine Journal of Science im 

order to free the oil, and this imparts the burnt odor and taste. 
The centrifugal cream separator suggests itself as a means of 
removing the water, and while it separates the water very nicely 
on a small scale, the solid particles of the emulsion rapidly pack 
into the separating bowl, and the operation must be interrupted 
while the accumulated fiber is removed. A modified centrifuge 
might be constructed that would not be subject to this disad- 
vantage. 

It was noticed that a large part of the water separated from 
the emulsion upon standing, and if allowed to stand for a 
sufficient length of time, microorganisms attack the emulsion, 
breaking down the tissue with liberation of the oil. The length 
of time required, together with the high rancidity and acidity 
of the oil produced, does not commend the process. Attempts 
were made to liberate the oil from the emulsion by heating 
with (1) salt, (2) acids, and (3) alkalies, but in all cases the 
oil was impaired in quality and the separation was not facilitated. 
Freshly grated coconut meat was heated in an autoclave at from 
120-130° C. for as long as four hours, both with and without 
the addition of water, but did not permit of a greater degree of 
expression than where live steam was used under ordinary pres- 
sure. Similar experiments were made, using an amount of 10 
per cent sodium sulphite solution equal to the volume of the 
meat on the assumption that the coconut meat would be disin- 
tegrated as in the making of paper from wood pulp with the 
sulphite process. This treatment might be used for the manu- 
facture of soap stock oils, as it facilitates the removal of the oil, 
but the quality of the oil is impaired in both color and taste. 

As a result of these preliminary experiments we have deter- 
mined methods that work successfully in laboratory practice and 
contain no points which could not be duplicated on a factory 
scale. The process which was found to yield the best results is 
outlined as follows : 

The meat may be removed from the shells and ground into 
a fine state of division by either of two methods: 

a. The process employed in the small oil mills throughout the 
provinces, which consists in grating out the meat from the half 
nuts by means of a revolving burr, manipulated . with a pedal 
attachment. One man can grate the meat at the rate of 100 
nuts per hour. 

b. A more modern method for dealing with large quantities 
of nuts divides the process into two parts. The meat is first 
freed from the shell by treating the nuts with live steam for a 
period of from fifteen to thirty minutes. It is convenient to use 



xii, a. 2 Parker and Brill: Pure Coconut OH 91 

a large hopper-shaped box provided with pipes carrying waste 
steam from the boilers. The treated nuts are opened, and the 
meat is readily removed with an ordinary copra knife, after 
which the meat is ground by the machinery employed for grinding 
copra. The meat prepared by the latter method is in a more 
finely divided condition than is obtained by the former method. 
This ground meat is next placed in a cooking vat provided with 
open pipes for live steam and mechanical beating apparatus, 
mixed with an equal volume of water, heated, and violently agi- 
tated for a period of three hours. By this treatment the oil 
is washed from the meat, together with some finely divided, 
cellular tissue which forms a white emulsion with the water. 
After the cooking process the emulsion is strained off and the 
pulp is passed into a molding machine, where it is pressed into 
cakes of convenient size for the final pressing by the hydraulic. 
Press cake so formed would necessarily require drying in order 
to improve its keeping qualities, since otherwise it would be 
attacked by microorganisms. The oil remaining could then be 
removed by further pressing just as in the case of any other 
copra cake. 

The emulsion resulting from the pressing, together with the 
strained portion, is now run into a storage tank, where it is cooled 
to a temperature of 15° C. and the temperature maintained for 
twenty-four hours. The freezing process allows the water to 
separate from the emulsion, leaving a supernatant layer of solid 
fat and cellular tissue, at the same time preventing the action 
of destructive microorganisms. After drawing off the water, the 
solid layer is allowed to melt by application of gentle heat or 
simply by standing, until it attains normal Philippine atmosphere 
temperature. The freezing and melting apparently rupture the 
cell walls and allow the oil to separate. The separated oil is 
next run through a filter press which removes "all solid materials 
present, then sterilized at 100° C. for thirty minutes, and stored 
in air-tight containers. Oil prepared by this method is water 
white in appearance, possesses a bland coconut odor, and is free 
from acidity and rancidity. The keeping qualities, even without 
sterilization, are superior to that from ordinary copra oil. A 
sample after standing for nine months showed only 0.6 per cent 
acid (oleic). 

A second method for the production of pure coconut oil and 
a valuable press cake from fresh nuts is outlined below. This 
does not eliminate drying, but is a continuous process in which 
the meat is removed from the shells and ground as described 
under the first method (either a or b), after which it is dried 



92 



The Philippine Journal of Science 



and, while still hot, subjected to pressure for the removal of the 
oil. In the finely divided condition the meat will dry in a current 
of hot air much more rapidly than when in large pieces, as is 
practiced in the making of copra. The preparation of shredded 
coconut, which is carried out by driving air at a temperature of 
70°-85° C. over finely divided meat spread on trays, requires 
only from thirty to sixty minutes for drying to a water content 
of 5 per cent. 2 

In a small experimental drier devised by us, the meat, after 
grinding, was dried to 10 per cent moisture content within an 
hour's time. This percentage of moisture is sufficiently low to 
prevent the formation of an emulsion. When the meat was 
expressed in a small horizontal press, over 80 per cent of the 
total oil was removed in one operation. 

Table II. — Data concerning the pressing of freshly grated and partially 
dried coconut meat. Pressure, 70 kilograms per square centimeter 
(1,000 pounds per square inch). Temperature, 80 to 85° C. 



l 
2 

3 

4. 



Min. 



P.ct. 

24.5 

36 

33.2 



P.ct. 

31.4 
9.0 
16.4. 
33.6 
25.6 
12.3 
9.0 



Extrac- 
tion of 
total 


Oil in 
cake. 


oil. 




P.et. 


P.ct. 


84 


31.8 


87 


24.7 


87 


29.9 


70.6 


44 


76.4 


27.4 


86.3 


22.6 


84.8 


24.6 


S9.4 


22.5 



Much. 
None. 
Some. 
Much. 
Little. 

Practically none. 
None. 
Do. 



The data on the extraction are valuable for comparative pur- 
poses, since they give the efficiency of expression of copra with 
varying degrees of moisture under identical conditions of ex- 
pression. A more efficient press would obtain more of the oil, 
but the ratio of yields with different moisture contents of cake 
would remain much as they are above. The point of interest is 
the fact that clear, water-white oil, free from emulsion, can be 
obtained from cake with a 12 per cent water content. 

A type of drier which should be efficient could be constructed 
along the lines of a sugar drier, or a drier constructed on the 
plan of that manufactured by the Philadelphia Textile Company 
would be effective in drying this finely divided meat. 



2 For further data on drying coconut meat, see Walker, This Journal 
(1906), 1, 147. 



xii. a, 2 Parker and Brill: Pure Coconut Oil 93 

In this process the meat would be machine-handled from the 
time it is removed from the shells until the oil and press cake 
are reached. We estimate that one expeller would be sufficient 
for a plant handling 45,000 nuts daily, giving an oil output of 
6 tons. The market quotation s for commercial oil on the Pa- 
cific coast at present is 27£ centavos (13jj- cents United States 
currency) per pound, and for Cochin oil the best grade of un- 
refined coconut oil on the market would be 35 centavos (17.5 
cents United States currency) per pound. The oil obtained by the 
process outlined above is water white and is free from acidity 
and rancidity; it would command the best market prices. This 
would mean an increase of 155 pesos per ton over the ordinary 
commercial oil, the grade to which Philippine oil at present 
belongs. 

Commercial copra press cake finds a place on the market in the 
form of stock foods and commercial fertilizers. As a stock 
food it compares favorably with gluten feed, but contains less 
carbohydrate material and more fat, ash, and fiber. According 
to data of the Massachusetts Agricultural Experiment Station 4 
100 pounds of coconut meal contains 88.4 therms of net avail- 
able energy as against 82.7 therms for the gluten feed, due to 
the higher percentage of fat of the former. (See same bulletin 
for more data.) 

The possibilities of the press cake obtained from the above 
method are worthy of consideration not only as a stock food, 
but for human consumption as well. This product is white and 
clean, containing a small amount of water and pure oil in amounts 
dependent, of course, upon the efficiency of the pressing. 

The keeping qualities would be better than in the case of 
shredded coconut, because of the lower oil content. Samples 
stored in closed and open containers for several months under- 
went no impairment in quality. Nutritive value and description 
of edible products, which we have prepared from samples of 
this meal, are as follows: 

Table III. — Nutritive value of copra meal from expression of freshly dried 
coconut meat. 

Per cent. 

Water 7.35 

Oil 32.14 

Ash 4.05 

Crude fiber 37.12 

Protein (NX6.25) 20.34 

'Oil, Paint & Drug Reporter (1916), 90, 63. 
'Bull. Mass. Agr. Exp. Sta. (1914), 155, 190. 



94 The Philippine Journal of Science 

Table IV. — Recipes for the preparation of food products from coconut meal. 



Coco- 1 Baking 

nut I Flour, pow- 
meal. der. 



Coconut oven 
scones. 

Coconut gems . 

Coconut grid- 
dle cakes. 

Coconut rusks*! 
Coconut bread.! 



'I '! 



Sug- 
ar. 


Eggs. 




Cup. 


Cup. 




i 


i 




i 


none 

1 

1 


1 


i 


1 





Salt. 



• One-half teaspoonful each of nutmeg and c 
sprinkled with susar and 



Sufficients 
make bat- 



until 
brown 
30-40 
30-40 



Manner of 
cooking. 



Fried in lard 

or crisco. 
Hot oven. 
Do. 



added to the dough, and the surface 



It will be noticed that in all of these products no lard, crisco, 
or butter is needed for shortening. Water may be substituted 
in all cases for sweet milk. In view of these laboratory tests 
it seems possible that coconut meal might be substituted to ad- 
vantage for wheat flour and lard in the preparation of cheap 
edible products. It must, however, be borne in mind that the 
meal used was obtained from pressing freshly dried coconut meat 
and not from ordinary commercial copra meal; meal from 
fresh machine-dried copra would be valuable for the same pur- 
pose. It seems to us that there is a good opportunity for a plant 
producing edible oil and press cake for human consumption. 

SUMMARY 

Methods used by the Filipinos, together with the results 
obtained by these methods, are presented. 

A method used by Mr. 0. Vyner, of British North Borneo, 
is reported. 

Experiments carried out by the Bureau of Science and their 
results are discussed. The description of a method, devised by 
us, which makes the grinding, drying, and pressing of coconut 
meat a continuous process and obviates the many handlings that 
at present are required, is included. 

Recipes for the preparation of food products from clean, pure 
coconut meat are given. 



THE RANCIDITY OF PHILIPPINE COCONUT OIL 1 

By Harvey C. Brill and Harrison 0. Parker 

(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

The cause, measurement, and prevention of rancidity in oils 
have engaged the attention of many investigators. 

Walker - decided that the rancidity of coconut oil arose from 
the action of heat, light, moisture, and air on the oils. He 
believes that most of the changes in the oil take place in copra 
due to the activity of the microorganisms and offers as a possi- 
ble explanation of the cause of rancidity in pure oils "that a 
small percentage of fatty acid is oxidized to an oxyacid, which 
in turn forms a lactone, and (assuming the formation of hydro- 
gen peroxide) the latter would give rise to a peracid, which, in 
turn, would oxidize the free glycerine to an aldehyde." Walker 
states that an oil which has been rancid and later refined again 
becomes rancid more readily than an oil that has never shown 
rancidity. No explanation for this greater tendency to rancid- 
ity is offered by him. 

As an estimation of the rancidity of coconut oil he determined 
the free fatty acids and made the fuchsin-sulphurous acid test 
for aldehydes and the peroxide test. He disclaims placing any 
credence on the free fatty acid determination as a measurement 
of the degree of rancidity of oils and asserts that the ordinary 
aldehydes are not offensively odorous. Therefore it appears 
from his results that 'the rancidity of oils must be due to 
the presence of compounds other than aldehydes or free fatty 
acids. Yet he maintains, "The most satisfactory [test for 
rancidity], in my experience, is that with fuchsin-sulphurous 
acids, shaking up about equal parts of oil and reagent." 

Haller and Lassieur 3 take issue with many other investigators 
regarding the cause of rancidity when they state that the un- 
pleasant odor of commercial coconut oil is due to the presence 
of free fatty acids, such as caproic, caprylic, capric, and lauric. 
In the volatile products from the refining of coconut oil these 
authors 4 found ketones. 

1 Received for publication March, 1917. 

2 Walker, H. S., This Journal (1906), 1, 1917. 

'Haller, A., and Lassieur, A., Compt. rend. Acad. sci. (1910), 150, 1013. 
4 Ibid. (1910), 151, 699. 

95 



96 The Philippine Journal of Science im 

Lewkowitsch 5 asserts— 

I therefore ascribe the primary cause of rancidity, namely the formation 
of free fatty acids, to the action of moisture in the presence of soluble 
ferments, which act as catalysts or accelerators ; * * * Rancidity is not 
due, as is still widely believed, to the presence of free fatty acids alone; in 
other words, rancidity must not be considered as coterminous with acidity 
* * * It is only when oxygen and light gain access to the acid fats that 
the conditions favoring the setting in of rancidity are provided. Rancidity 
is rather due to the direct oxidation of free fatty acids by the oxygen of 
the air, assisted and intensified by the exposure to light. 

The presence of free fatty acids can be, therefore, considered 
only the first step in the formation of rancidity. 

Batter placed in storage for prolonged periods of time often 
acquires a rancid odor. This rancidity has been ascribed to 
the activity of microorganisms by various authorities. The 
latest of these investigators is Dyer," who has shown by a thor- 
ough study of cold-storage butter that the change in flavor is 
due to the slow oxidation of the nonfatty portions of the butter. 

We believe that the nonfatty material in coconut oil has a 
profound influence on its character, and experiments are now 
in progress to test this out. 

Schmid 7 classifies oils as "sour fats," "rancid fats," and 
"sour and rancid fats." In coconut oils we have distinct classes 
of odors existing in the rancid oils, namely, sour odor, pos- 
sessed especially by some oils made by Filipinos from fresh 
nuts which are characterized by a high acid value ; 8 esterlike " 
and smokelike odors, observed particularly in oils made from 
moldy copra which has been prepared by the tapahan (Filipino 
grill for drying copra meat) method; and a third with an odor 
of stale lard, which is shown by sweet coconut oil that has be- 
come rancid due to exposure to the air. The investigation now 
under way will attempt to determine the difference in the chem- 
ical character of the various odors and whether the cause of 
each particular odor is peculiar to it. 

The present article deals with the various methods of meas- 

* Lewkowitsch, J., Chemical Technology and Analysis of Oils, Fats and 
Waxes. Macmillan & Co., London (1913), 1, 52. 

'Journ. Agr. Research (1916), 6, 927. 

1 Schmid, A., Zeitsckr. f. anal. Chem. (1910), 37, 301. 

8 For method of preparation see Parker, H. 0., and Brill, H. C, This 
Journal, Sec. A (1917), 12, 

Coconut meat which has been acted upon by bacteria possesses an 
esterlike odor, and this odor in coconut oil probably arises from the action 
of the bacteria on the meat before pressing. 



xii, a. 2 Brill and Parker: Philippine Coconut Oil 97 

uring rancidity which have been recommended by the inves- 
tigators of rancidity in oils. 

One of the methods, which has many advocates, as a test for 
rancidity is the fuchsin-sulphurous acid color test. Walker "' 
reports that in many commercial oils which gave organoleptic 
tests for rancidity the color test failed to respond, while in many 
instances oils which were sweet did give color tests. This test, 
which is known as Schiff' s aldehyde test, is considered by some 
to owe its activity to the presence of active oxygen for results. 
Aldehydes have the property of activating the oxygen of the air. 
One of the chemists to investigate the property of aldehydes to 
give oxygen was Ludwig. 11 He found that the aldehydes inves- 
tigated by him— acetaldehyde, paraldehyde, propylaldehyde, 
isobutyraldehyde, and benzaldehyde— lost this power of giving 
a reaction with potassium iodide if they were previously dis- 
tilled in a stream of carbon dioxide. Urbain 12 claims that the 
color, produced by aldehydes with decolorized fuchsin, is the 
result of a condensation reaction. 

Kastle and Loevenhart ' 3 ascribed the oxidation activity of 
benzaldehyde to the absorption of oxygen from the air to form 
benzoyl hydrogen peroxide (C K H s CO-0-OH). This compound 
readily gives up one oxygen and forms benzoic acid, which no 
longer possesses the property of activating the oxygen of the 
air. These authors believe the activity of oxidizing enzymes 
arises from a similar reaction. 

The fact that certain oils which are unmistakably rancid do 
not restore the color to decolorized fuchsin, while other oils which 
are not rancid do produce a color with the reagent, shows that 
the test is not reliable as an indication of the rancidity of the oil. 

Additional objections to the use of the fuchsin-reagent are that 
it gives a color reaction with compounds other than aldehydes 
(assuming that aldehydes are present when rancidity exists) 
and that these colors are difficult and often impossible to distin- 
guish in shade. 

An alkaline solution of diazobenzene sulphonic acid reacts with 
aldehydes in the presence of sodium amalgam to give a violet-red 
coloration. This reagent was used in the tests recorded in Ta- 
ble I. 

Table I gives the color reactions of fuchsin-sulphurous acid 

" Op. cit. 

"Ludwig, E., Ber. d. deutsch. chem. Ges. (1896), 29, 1454. 

u Urbain, G., Bull. Soc. chim. (1896), 15, 455. 

u Kastle, J. H., and Loevenhart, A. S., Am. Chem. Joum. (1901), 26, 539. 

149622 4 



98 The Philippine Journal of Science 1917 

and of diazobenzene sulphonic acid with a few organic acids, 
alcohols, and aldehydes. 

Table I. — Color reactions of organic acids, aldehydes, and alcohols. 



Name of compound. 


Result with fuchsin-sulphur- 
ous acid. 


Result with diazobenzene 
sulphonic acid. 


Formic acid 


None 


Slightly violet. 

Reddish violet. 

Yellow. 

Cherry red. 

Yellow. 

Slightly violet after short 

time. 
Reddish violet. 
Deep violet directly. 
None. 

Do. 
Deep violet after short time. 
Violet. 
Violet (good). 

Do. 

Do. 
None. 

Do. 

Do. 

Do. 
Faintly violet. 

Do. 

Do. 
Violet. 
Faintly violet. 

Do. 
None. 
Red. 
Yellow red. 


Acetic acid 




P • 'd 


















None 

Violet _ 






Oxalic acid - 


do 

do 


Free fatty acids from coconut oil » 


















Glycerol (old; slightly colored) 

Glycerol (old; air passed through 


do 















































* This coconut oil gave no test with the reagents before saponification. 

The compounds used for the color tests were Kahlbaum pro- 
ducts. A study of the results indicates that the tests are 
not due to the presence of aldehydes, since identical reactions 
occurred with other compounds. 14 Color reactions were given 
by acetic, propionic, butyric, valeric, stearic, and glycollic acids; 
by ethyl, butyl (normal), isobutyl, butyl (tertiary), amyl, heptyl, 
and ethylene alcohols; and by acetyacetone with the fuchsin- 
sulphurous acid reagent. Positive or confusing tests were given 
by acetic, butyric, stearic, oleic, glycollic, and malonic acids ; 
by ethyl, butyl (normal), isobutyl (tertiary), amyl (normal), 



14 A more thorough investigation of the property of organic compounds 
to give color with these reagents is now under way in this laboratory. 



xii. a, 2 Brill and Parker: Philippine Coconut Oil 99 

and ethylene alcohols; and by acetylacetone with diazobenzene 
sulphonic acid. Tests were given by the free fatty acids of 
coconut oil, by formaldehyde, acetaldehyde, and paraldehyde 
with both reagents, but the four samples of glycerol responded 
negatively. 

Color reactions were given by various alcohols, aldehydes, and 
acids with diazo-benzene sulphonic acid and with fuchsin- 
sulphurous acid. As these compounds are Kahlbaum chemically 
pure products, it hardly seems plausible to assert that the com- 
pounds reacting to give the color could be aldehydes present as 
impurities. 

In Table II are recorded some data on coconut oils in the pos- 
session of the Bureau of Science. 

An examination of the records of Table II reveals some in- 
teresting relationships, namely, Nos. 38 and 44, with low acid- 
ity, both show rancidity when tested by their organoleptic 
properties ; on the other hand, Nos. 13, 22, 26, 34, 36, 39, 46, and 
47 show relatively high acid values compared with that of fresh 
oil (usually about 0.2 per cent calculated as oleic), but are free 
from rancid odors. These results indicate that the acidity of 
coconut oil is not a measurement of the rancidity. 

Samples 2, 5, 7, 8, 9, 10, 11, 32, 37, 38, and 45 have a rancid 
odor, but do not restore the color to decolorized fuchsin, while 
the color is restored by samples 26, 28, 33, 40, and 41, which 
do not possess a rancid odor. The existence of no relationship 
between the restoration of the color of fuchsin and the odor of 
the oil is indicated by these results. The results with diazo- 
benzene sulphonic acid are equally interesting. Oils with rancid 
odors which do not give a color are Nos. 1, 3, 8, 9, 10, 38, and 45, 
and those giving a color but not possessing a rancid odor are Nos. 
26, 28, 36, 39, 40, 41, and 42. 

If the color with these two reagents arises from the same 
cause, it is impossible to explain the formation of color products 
with decolorized fuchsin in the case of samples 1, 3, and 33 and 
the absence of color of these samples with diazobenzene sulphonic 
acid, or the color given by Nos. 2, 5, 7, 11, 32, 36, 37, 39, and 42 
with diazobenzene sulphonic acid and the absence of color with 
decolorized fuchsin. Diazobenzene sulphonic acid is considered 
a more sensitive reagent for the detection of aldehydes than is 
fuchsin. This explains why color resulted with diazobenzene 
sulphonic acid, but not with fuchsin, the latter results, but does 
not explain the preceding results. Oils which agree in all re- 
spects in the so-called rancidity tests are Nos. 17, 18, 19, 
and 27, which possess no rancid odors, no high acidity, and do 



100 



The Philippine Journal of Science 



■5% 

is 


None. 

Salmon immediately. 

None. 

Pink after short 

standing. 
Faint violet imme- 
diately. 
Do. 
Yellowish red imme- 
diately. 
None. 
Do. 
Do. 

Yellowish immedi- 
ately. 

Medium violet on 
standing. 

None. 

Medium violet im- 
mediately. 

Very faint violet on 
long standing. 

None. 


~3 

•E-o 

■sH 

3 00 

if 

o 

a 


H 

C 

'5 

3 

2 6 
.2 .y 
> -5 

u c 

. o I 


"o 




< 

1 

1 


s 
1 


o 

•a -c 


: : i 

: : E 

| 

I : ° 
• J . 
•3 h 

* . 

■0 -o « a « 
! ! "5 -5 a 
! ! ,5 fc 


i 5 ' 

e x : 

1 : ; 

1 ! 

E £ 
* — >> 
,tc a x . ■ 
•5 tj j* 3 
£ S3 -55 


=5 «~T3 


cC 


"o 


41 

! 1 

2 £ 

PL. cc 


"o 
■s c 


"a 
- t 

5 


c 


\ 


o 

I 1 

ll 


i ; 1 
s 1 1 

"3 ss "3 

1 1 1 4! 
S 2 S 


| 

« 

1 * 


o 
13 
O 


< 
I 


e 
* 


o c 
■a t 


1 

E 
u 


^ 






e 


| 

e 
Of 


= 

i S 


1 
I 


•J CO 


S°« Si 
<£E« 


5 


I 

? 

o 

ft 

Q 


■e 

a 

i 

j 

i 


* 


o e 
■a x 


c 


c 


e 


5 

c 
a 

V 

'£ 
a 

s 

'i 

- c 
( 

1 

It 

p- 


c 
"0 


•a 

el 

I .<• 

ai : 

." ? S 

"5 1 2 

P 2 1 

xi • < 

2 2 - 

■» r r J! 

a *i x 

w 8 x 
w j 

? ? j 
| 1 1 

C C 

g f. g 1 

< « i 

Pm a< 2 


4 


£ 7S 

1 1 3 

! g 

: I 1 I 

1 ! P 

-v 1 S- 
1 1 -s 2 . 

S g ? "S 

E E & I 

Ii.fi 

* * 2 * 
5 5 5 


o 
1° 


•h cm eo ■>» ia w t- ooeso jh eoj 3 3 S 2 



xii, a, 2 Brill and Parker: Philippine Coconut Oil 



101 



a a 



o o 5 « "o a o « — "a 

> Z fa q fa iz; > co 

!i I I I. i i ! !•:!!! 

II ! ~ I a I j ! i ! i I : 

- 1 ! I s J j I | i \ I I i 






I ; I ! j 1 j I j | j i • | j | 

i i i i.i 1 i | il ■! I i i | j 

£ n I £ I [ I ; n 






1 1 



o ? ; 



I i 

It III 

§ * 2 2 2 

„ -g 2 * c n 

$ > ? l a * i s s i ; 8 

S S 3 |§ | -g = I "S 1 ^ 

gkP^" 1 " 18 ^ E-F- 2 ° 

it'-ll--! 51*11 

65 55hooE 

s s s s a s s s 



« .2 t rt x o 
5 A -" S te 1 



- - 1 * -S 

3 3 | "g g 



£ £ ~ "S "S 

fa fa o 



2 .5 



s*3 
"3 o 



c <* • 2 



s jj a a § o. 

fa fa (J 



102 



The Philippine Journal of Science 



\l 

~ £ 

•O o 

■5 o. 

|a 

as 


None. 

Do. 
Faint violet soon. 
Very faint violet on 
long standing. 

Do. 

None. 

Very faint violet on 

long standing. 
Faint violet soon. 

Do. 

Faint violet. 

Deep violet imme- 
diately. 
Violet immediately. 

N 
None. 

Do. 


"3 
•E-d 

J= o 
_o 

6 


1 
c 
'3 

i i 

> 2 


' o ; ! 

,6f j 

•5 ! 

.£ ! i 

be T3 -C 

S % I 
> 2 ! 


: : : i \ 
: ; : : e j* 

; £ • 1 £ 
i "S i i - i J 

•° c ^ g .5? .« .2 « 

"3 o -a T3 .2 I 
| fa j 55 .5 > 2 


i 


2 *'5~' 


** o oi »£> ci <m* o rA © o o ■* o 

a,' 


t 


j | 

6 


= 3 

I 1 

fa hJ 


! J 
• : 

3 1 

: ™ 


| 


o 
1 

fa 





! i i : i * i 

1 ' — ! 
! ! ! ! ! £ ■ 

•a -o •§ ■§•§£•§ 

! ! : : t« : 
; | 3 j 


•a 
O 




E § 


9 


i m 

: * 


1 I 1 ! a 

! : ° i £ 

o o o -so - ; 

■"5 "°. "°, "2 ^ o 

: : ; « : e 

... h3 ; to 


1 

is 
w 


Ap- 

proxi- 
mate 
age. 


1 








2 

I 
1 

3 
O 

.s 
c 

1 

■5 
E 

•s'l 

5 


c. 
£ 
a 

J 
& 

O 


T3 

o. 
a 
2 

1 1 
fa 5 

£ 1 
1 | 

j I 

'S ^ 

"8 | « 

lis 

E £ 


j c 

! Q 
; £ 

! a 

— 
3 '5 

ll 

"3 61 

II 

O 3 

O fa 


Pure Lugga olive oil from Bureau of Science 
storeroom. 

Cotton seed oil from Bureau of Science store- 
room. 

Castor oil from Bureau of Science storeroom 


Unfiltered oil from grated meat, freshly dried 
and pressed (Feb. 16. 1917). 

Filtered oil from grated meat, freshly dried 
and pressed (Jan. 20. 1917). 

Oil from grated meat, freshly dried and 
pressed, filtered three times through Ful- 
ler's earth, not sterilized; in open container. 

Oil from machine-dried copra (copra one and 
one-half years old). 

Oil from grated meat, freshly dried and press- 
ed (air space over oil). 

Oil 32, neutralized with NaOH, washed and 




"mOTMeoS§CT°5^ 3 3 » S 



xii, a, 2 Brill and Parker: Philippine Coconut Oil 



103 



o 



it J 

I 1 



"Si 



♦3 & 

*? 2 



104 The Philippine Journal of Science isn 

not give any color with either reagent, and Nos. 4, 12, 20, 21, 23, 
24, 25, 29, 30, 31, 35, 43, 44, and 48, which show color with 
both reagents and possess a high acidity and rancid odor. In a 
total of forty-eight oils with four tests for rancidity being used, 
one or more of these tests do not agree with the others in 
thirty cases. 

The organoleptic properties are the most dependable tests for 
rancidity. In Table III the relationships of the other three tests 
to the organoleptic properties are given. 

Table III. — Relationships of rancidity test to the organoleptic properties. 



Test. 


Disagree- 
ments. 


Agree- 
ments. 




Per cent. 
21 
29 
33 


Per cent. 
79 
71 
67 









From Table III it appears that high acidity is a more 
reliable indication of the rancidity of an oil than the reactions 
with diazobenzene sulphonic acid or decolorized fuchsin, but this 
does not prove that rancidity is coterminous with acidity. They 
occur simultaneously in many cases, but the one is not a neces- 
sary corollary of the other. When oils of high acidity are steam- 
distilled, the residue no longer possesses a rancid odor, but has 
lost very little in total acidity. Samples 46 and 47 bring this 
out very prominently. The acidity in these samples remains 
high, but the rancidity, judged by the odor, has been lost. 
Sample 45, which had been neutralized with sodium hydroxide 
and carefully washed, was not freed of its odor though freed 
of its acidity. Sample 48, which was washed with a solution of 
sodium hydrogen sulphite to rid it of any aldehydes that might 
be present, still retained its rancid odor. If rancid odor is due 
to the presence of aldehydes, this sample should have lost its odor. 

Issoglio 15 describes a confirmatory test for rancidity in olive 
oils, which he calls the "oxidisability value." He states that 
when the oxidisability value exceeds 15 it will be usually found 
that the oil is rancid or has undergone some other change. 

Table IV contains some oxidisability values and some acetyl 
values, together with other data on Philippine coconut oils 
before and after distillation with live steam. 

"Issoglio, G., Ann. chim. applic. (1916), 6, 1. 



a. 2 Brill and Parker: Philippine Coconut Oil 



105 





g!H 








*J « 




o « 


_o j» _« 




8| 


!s 33 


*j 


=sg 


j > „• > '? j 


c 

J 
| 


*1 


| c 1 c c -1 


o « 


g O c * g c £ « & £ gO&S«OOc 


•s c 


"3 v '5 « « '5 


g 


6S 


Z fa > Z h Z > Z > Z fa 
















| 


s 














§ 


!•- 












! : : ! ! : i ; ! j j 


o 


■o.c 




i % \ 






!!!:!:!!!!: 


S 


.■s«- 




> 3 > 








| 


*s 






• J 


00000000000 


2 


fe.2 


8 "o e "° g« •mflflfltij'O'Otij'a 


5 


u 


! "5 1 .£ ! ! ! ! ! ! ! ! ! 

z:fo:z>z:::::;::: 


1 


>. « 


~ » * n . 1 «• « «- 1 N . 1 » oi -! »* i i 






s-sl 


"OOOOCOCOOOCCOCoOOO 

b; c g g g g •? <j « ^ 




c-d 


















:;;:::: 




•S'c 














' ' « 




i \ \ 1 1 1 ! 
















: ; 




3 > ' ! ! ! I 


1 


■5 o 
£ a 
'go 


■3 
|3 

c.S 
•5 J 


II 


|3 

■3 .S 


i .5 

■§ c >> % 
! '3 S 3 


.s i : & ; ; 

e & c •§ | & | ■§ 
'5 id a 1 .2 .« .2 ! 


i 


6S 


fejHfa>ix,> ; fa > > fa > fa ;>>;> ; 


J 


















i 




















1 


g c 








! ! ! 






! ! ! ! 1 


""■S 






1 .. 3 






_ 


5 3 




! 


! a 3 1 .' .»f 


.jo 


! 2 ■' 


o 








; "3 j- ; ; -o 


•5 




"3 

3 
■V 


*11 

o — 

■3 




' > « 1 = 1 • ai 

! "2 > 'S S "2 ! '3 '5 § > £ ■ : ; 2 

z : & : £ <3 .5 ifaZhZ ; : : ; 8 


>>.2 


^jC0O«DCJ0a>O'd'0000tX)N0aOONN*^e0 




J5l3 


» ti « ei ti d » h h ^ «' el h 10' si oi e> m 0' 




Is- 


ft] 


« M 


^iacoict<ooou]aa)ooNNHNNni04 


Acidi 
of ori 
inal 
oil ai 
olei< 
acid 


ti tj «' ei d t- « ri » e»' d e>i 10 n ei d h ci 

a," 


° « 


Ol ID O O) IO 


» H S >» B 


10 







•S 3 

51 


cd c^ id uS ^ 


to od i> cd 10 


CO 




<d 














*?2 


iafa)(0»«iowoiobiD(Ow»OHt- 


si 


d ^ d d d d m ^ t^ d d d d d h « d « 




<> 


rH 


3 = — 
'S2« 


ocoooooa»u5c^ocotoot-cntDTj"o 
h id t-' d w d d h d h d c^ d d ci ^' ^' d 


0« * 




















"5 












1. 










a 


e00oo2ioooo2e.!og!;e 
o-otstj-oC g t) ■» tj ■» fl § « ns 3 


M 


; ; ; |~o ; ; ! ; 5 o — ; o « o 
O 1 J ; j fa O | i j j o< uuu 




•o -° 




'5,*i 




s°s 


+ + + +I++ + + +I I + I++I + 


" to 




a". 


■'^SSSSSmcoOTcocoeOMmmm^ 


E ° 




§z 




ra 





s- S 



106 The Philippine Journal of Science 1917 

The oxidizability value for the coconut oils exceeded 15 (the 
maximum value given by sweet olive oils according to Issoglio) 
in samples 4, 11, 14, 32, 35, and 44. All of these samples were 
rancid, but samples 15, 30, 31, and 37, which were rancid, gave 
values lower than 15. It is possible that a limiting value of 
15 for coconut oil may be too high, but the limit cannot be 
brought low enough to include all the rancid oils examined by 
us and still exclude the sweet oils. The test appears to be a 
good confirmatory one, since no sweet oils gave high results. 

The acetyl values of several oils were determined in an endeav- 
or to ascertain if any relationship exists between them and the 
rancidity. The lowest value, 7.0, was given by No. 31, a rancid 
oil, while the highest value, 21.7, was shown by No. 11, also a 
rancid oil. The value of the acetyl number is dependent on 
the presence of the hydroxyl group. Lewkowitsch makes no 
mention of the presence of hydroxy acids in coconut oil. The 
only unsaturated acid generally accepted as present in coconut 
oil is oleic, which exists in comparatively small quantities. It 
is possible that when rancidity takes place one of the reactions 
occurring is the formation of oxy or dioxy stearic acid from 
oleic acid, but this would not account for an acetyl number of 
a magnitude of 21.7 for coconut oil. The residue of the increase 
must arise from the freeing of the hydroxyl groups of the 
glyceryl radical. A high acetyl number should be, therefore, 
associated with a high acid value in the case of coconut oil. A 
study of the contents of Table IV proves that this relationship 
does not exist in all instances. It is possible that where the 
hydrolysis of the oil takes place in the copra itself much or all 
of the free glycerol remains in the press cake or is further de- 
composed, resulting in a low acetyl number along with a high 
acid number. 

The iodine number of an oil with a high acetyl number should 
be correspondingly low if the increased value of the acetyl nunv 
ber arises from the oxidation of the unsaturated acids. The data 
in Table IV do not prove the validity of this assumption, since 
no definite relationship appears to exist between the values of 
these two chemical constants. 

Further tests were made on four oils. These data are re- 
corded in Table V. 



Brill and Parker: Philippine Coconut Oil 



107 



c a Q o 



z £ 



O r-c rf 



S 6 
c . 
•I E* 



s g 



-. O O 



5 b 



108 The Philippine Journal of Science xm 

Sample II was sour in odor, No. 19 was a sweet oil, No. 32 
was smokelike in odor, while No. 50 was lardlike in odor. 

The iodine values of the steam-treated oils are lower in all 
four cases than are the iodine values of the original oils. Ap- 
parently the unsaturated compounds are either volatile in the 
steam or they are partially changed to saturated compounds by 
the live steam. 

The acetyl values in three cases are lower for the treated oils. 
The hydroxy compounds appear to be somewhat soluble in the 
live steam. The higher value in the case of No. 11 probably 
arises from the hydrolysis of the glyceryl esters of the fatty acids 
by the live steam. The free fatty acids which are present in com- 
paratively large quantities in this sample act as accelerators of 
this hydrolysis of the neutral oil, or the mono- and diglycerides 
may be further hydrolized to free fatty acid and glycerol and 
to the monoglyceride and glycerol, respectively. The Reichert 
Meissl number shows the variation one would expect with the 
steam treatment of the oils. No difference in its value was 
found for the sweet and the rancid oils. 

The free fatty acids and their relationship to the rancidity 
have been already discussed. This constant is included here 
in order that the acid value for No. 50 might be tabulated. 

The relative values of the soluble fatty acids vary for the orig- 
inal and treated oils in a manner difficult to explain. Further 
data will be necessary before any conclusions can be drawn 
from this constant in regard to the effect of steam distillation. 
Dakin 16 has shown that when oleic acid is oxidized with hy- 
drogen peroxide azelaic acid is formed and all fatty acids 
from formic up to and including stearic are oxidized by mild 
oxidizing agents. If the formation of rancidity is caused or 
accompanied by an oxidation, one of the first changes to take 
place would be the breaking down of the unsaturated acids into 
simpler acids with an accompanying increase in the value of 
the soluble fatty acids. Consequently the value of the soluble 
fatty acids of rancid oils should be greater than the value for 
the sweet oils. This generalization holds true for the original 
oils that have been tabulated in Table V. 

The milk from the coconut contains both oxidase- and perox- 
idaselike enzymes. These enzymes are not found in the pure, 
fresh coconut meat, and their presence in such meat is probably 
due to contamination from the milk of the meat. However, 

"Dakin, H. D., Journ. Biol. Chem. (1908), 4, 63 and 237. 



xii, a, 2 Brill and Parker: Philippine Coconut Oil 109 

copra which has been attacked by molds or copra with milk 
adhering to it has oxidasing enzymes present, and when such 
copra is ground and pressed, some of this enzyme may be sep- 
arated with the oil. If such enzymes were present, they would 
in all probability have an important influence on the formation 
of rancidity. 

The oils described in this article were examined for oxidizing 
enzymes. No. 5 gave a faint test, No. 15 gave a fair test, and 
No. 28 gave a strong test for peroxidaselike enzymes with 
tincture of gum guaiacum and hydrogen peroxide. All the rest 
responded negatively for oxidizing enzymes with these reagents. 
No conclusion can be drawn concerning the cause for the pres- 
ence of the enzymes in these samples and their absence in all 
the others. No. 5 is a commercial sample of oil with a smoke- 
like odor, which indicates that it was made from tapahan-dried 
copra. In method of preparation and history it is very similar 
to Nos. 1, 2, 3, 4, 5, and 7. No. 15 is an oil made from a machine- 
dried copra. This sample is very similar to No. 14 in history. 
No. 28, made from fresh copra, is pure and sweet and similar 
in most respects to No. 27. The enzyme here must arise from 
contamination from the milk of the nut. The activity of the 
other two must arise from the molds that grew on the copra 
before it was milled, the milling temperature not being high 
enough to destroy their activity. 

Why only these three gave positive tests is difficult of expla- 
nation. The color appeared in them only after the sample had 
stood in contact with the reagent for a short time. 

Tests were made on the fresh coconut meat for lipase, but 
we were unable to confirm the positive results reported by other 
investigators. 17 Walker 18 reports no lipase or zymogen pres- 
ent in the fresh meat, and our results are in accord with his. 
Where the meat of the coconut is subjected to a comparatively 
high temperature during the process of drying, the activity of 
any enzyme doubtless would be greatly decreased if not totally 
destroyed. But the hydrolysis of oils in the presence of mois- 
ture undoubtedly takes place to an appreciable extent at rela- 
tively low temperatures. In the experiments recorded by 
Deming, where comparatively large amounts of mineral acids 
were used "to activate the zymogen," we believe the acid alone 
is sufficient to bring about the hydrolysis recorded by this observ- 

" De Kruff, Bull. Dept. Agr. Indes Neerl. (1906), 4, 8. Deming; H. G., 
Phil. Agr. For. (1914), 3, 33. 

"Walker, H. S., This Journal, Sec. A (1908), 3, 111. 



HO The Philippine Journal of Science 

er. Acetic acid failed to give like results when used by us, 
and were a zymogen present this acid should activate it. 

CONCLUSIONS 

The color tests with decolorized fuchsin and with diazobenzene 
sulphonic acid are not reliable tests for rancidity. High acid- 
ity of oils is not coterminous with rancid ty. Steam distillation 
removes rancidity, but makes very slight changes in the acidity; 
neutralization with alkali and washing removes the acidity but 
not the rancidity. The Reichert Meissl number in the few cases 
studied indicated no close relationship between this constant 
and the rancidity. No conclusions could be drawn from the 
magnitude of the iodine value. The soluble fatty acids show 
a relationship which indicates that this constant might be of 
value in an estimation of the rancidity of an oil. The acetyl 
value may be of value in indicating rancidity, but is not a meas- 
ure of the degree of rancidity. The oxidizability number ap- 
pears to be a good confirmatory test for rancidity, but a few 
samples of oils undoubtedly rancid gave low oxidizability values. 
However, where the value was high, the oil was always rancid. 

Further work on the acetyl value, the oxidizability Value, the 
soluble fatty acids, and the test for the presence of oxidizing 
enzymes and their action should yield valuable data. 

Control samples have been stored for the study of these 
constants. 



PUBLICATIONS FOE SALE BY THE BTTBEAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS — Continued 



BOTAOT 

A FLORA OF If ANILA 
By Elmer D. lusuuu. 

Order No. 419. Paper, 490 pages, $2.50, 

postpaid. 
Practically a complete flora of the cul- 
tivated areas In the Philippines. Descrip- 
tions, with keys, of over 1,000 speoles, 590 
genera, and 136 families, with native names, 
glossary of technical terms, etc. 



PHILIPPINE DIFTEROCARP FORESTS 

By William H. Brown and Donald M. 

Matthews 

Order No. 432. Paper, 150 pages, 1 map, 
13 plates, and 12 diagrams, $1.00, 
postpaid. 
fn Philippine Dlpterooarp Forests the 
authors present a. very comprehensive discus- 
sion of the growth and development of dip- 
terocarp trees and of the Other elements of 
lowland Philippine forests. 



IND0-HALA7AN "WOODS 
By Fred W. Foxworthy. 

Order No. 411. Paper, 1S2 pages, 9 

plates, $0.50, postpaid. 
In Indo-Malayan Woods, Doctor Fox- 
worthy has brought together a large amount 
of accurate information concerning trees 
yielding woods of economic value. 



ZOOLOGY— Continued 

A MANUAL OF PHILIPPINE BIRDS 

By Richard C. McGregor 

Order No. 103. Paper. 2 parts, 769 

pages, $4, postpaid. 

A Manual of Philippine Birds contains 
in oompact form descriptions of all the 
known species of Philippine birds. The usual 
keys and diagnoses of orders, families, and 
genera help the novice In identification. 



By David Starr Jordan and Robebt Earl 
Richardson 



This list will be found a convenient guide 
to the synonymy of Philippine Ichthyology. 
The nomenclature is thoroughly revised, and 
the distribution of each species within the 
Philippine Islands is given. 



By W. Schultzh 



A LIST OF. MAMMALS OF THE 
PHILIPPINE ISLANDS, EXCLU- 
SIVE OF THE CETACEA 

By Ned Holusteb 

Order No, 418. Paper, 64 pages, $0.50, 

postpaid. 

The distribution of each speoies Is given, 
and the original descriptions are cited. 



Paper, 19S pages, $1.1 



Order No. 436. 
postpaid. 

This catalogue Includes the names of all 
species of Coleoptera that have been recorded 
from a definite locality in the Philippine 
Islands. References to original descriptions 
and other Important notes are given. The 
economio appendix Includes comment on 
those species of beetles which are known to 
be injurious or beneficlaj to man. 



PRICES ARE IN UNITED STATES CURRENCY 

Orders for these publications may be sent to the BUSINESS MANAGER, 
PHILIPPINE JOURNAL OF SCIENCE, BUREAU OF SCIENCE, MANILA, P. I., 
or to any of the agents listed below. Please give order number. 



The Macmillan Company, 64— 66 Fifth Avenue, Xew York, U. S. A. 

Win. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinns iNijhoff, Lange VoorhOut 9, The Hague, Holland. 

Slayer & Miiiler, Prinz Louis Ferdlnandstraese 2, Berlin N. W., Germany. 

Kelly & Walsh, Ltd., 32 Raffles Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 19 Bui Hie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



CONTENTS 

COX, ALVIN J. The study of copra and other coconut products.. 

BRILL, HARVEY C; PARKER, HARRISON O.; and YATES, 

HARRY S. Copra and coconut oil 55 

PARKER, HARRISON 0-, and BRILL, HARVEY C. Methods 

for the production of pui j coconut oil... 87 

BRILL, HARVEY C, and PARKER, HARRISON O, The ran- 
cidity of Philippine coconut oil . ... 95 



u. s. 

The "Philippine Journal of Science" is issued as follows: currency. 

Section A. Chemical and Geological Sciences and the Industries- $2.00 

Section B. Tropical Medicine - 3.00 

Section C. Botany 2.00 

Section D. General Biology, Ethnology, and Anthropology (Sec- 
tion D began with Volume V) - ----- 2.00 

Entire Journal, Volume II, III, IV, or V 5.00 

Entire Journal, beginning with Volume VI - - 7.00 

Single numbers (except of Volume I) — - .50 

Each section is separately paged and indexed. 
Authors receive 100 copies of their papers free. 
Volume I, 1906 (not divided into sections) and supplement, sold 

only with a complete file of section A, B, or C 10.00 

Supplement to Volume I (botany) 3,50 

Volume I (without supplement) , sold only with a complete file of 

section A, B, or C— . - , 6.50 

Single numbers of Volume I .75 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, Pi, I., or to any of the agents 
listed below: 

AGENTS 

The Mapmillan Company, 64— 66 Fifth Avenue, New Vork City, TT. S» A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinus Nijboff, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdinand strasse 2, Berlin, N. W., Germany. 

Kelly & Walsh, Limited, 32 Baffles Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 1 9 Baillie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



ed at the post office at Manila, P. I., as second-class matter 



May, 1917 



THE PHILIPPINE 

JOURNAL, OF SCIENCE 



ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OP 

H. C. BRILL, Ph. D.; J. R. WRIGHT, Ph. I).; G. W. HEISE, M. S. 
J. C. WITT, PH. D. ; T. DAR JUAN, A. B. ; A. H. WELLS, A. B. 

r. c. McGregor, a. b.;.h. e. kupper, a. b. 




MANILA 
BUREAU OP PRINTING 

1917 



PUBLICATIONS FOE SALE BY THE BITEEAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



ETHNOLOGY 

A VOOABULABY OF THE IGOROT LAN. 

(WAGE AS SPOKEN BY THE 

BONTOC IGOBOTS 

By Walter Clayton Clapp 

Order No. 408. Paper, 89 pages, $0.75, 

postpaid. 
The vocabulary is given In Igorot-English 
and Enolish-lgorot. 



THE NABALOI DIALECT 

Br Orto Scheerer 

and 

THE BATAK8 OF PALAWAN 

By Edwahd Y. Miller 

Order No. 403. Paper, $0.25; half mo- 
rocco, $0.75; postpaid. 
The Nabaloi Oialeot (65 pages, 29 
plates) and the Bataks of Palawan (7 
pages, 6 plates) are bound under one cover. 



THE BATAN DIALECT AS A MEMBEH 

OF THE PHILIPPINE GROUP 

OF LANGUAGES 

By Otto Scheerer 

and 

"F" AND "V" IN PHTZIPPIHK 
LANGUAGES 

By Carlos Everett Conant 

Order No. 407. 

These two papers are Issued under on* 
oover, 141 pages, paper, $0.80, postpaid. 



ETHNOLOGY — Continued 

STUDIES IN HOBO HISTORY, LAW. 

AND RELIGION 

By Najeeb M. Saleebt 

Order No. 405. Paper, 107 pages, 16 
plates, 5 diagrams, $0.25; half mo- 
rocoo, $0.75; postpaid. 
This volume deals with the earliest 

written records of the Moros In Mindanao. 

The names of the rulers of Maglndanao are 

recorded In five folding diagrams. 



NEGBITOS OF ZAHBALES 

By WrLLiAM Allan Reed 

Order No. 402. Paper, 83 pages, 62 
plate*, $0.25; half morooeo, $0.75; 
postpaid. 
Plates from photographs, many of which 
were taken for this publication, show orna- 
ments, houses, men making flra with bamboo, 
tows and arrows, dances, and various type* 
of the people themselves. 



INDUSTRIES 

PHILIPPINE HATS 

By C. B. RoBmsoN 

Order No. 415. Paper, 66 page*, 8 
Plates, $0.50 postpaid. 

This paper Is a concise record of the 
history and present condition of hat making 
In tha Philippine Islands. 



THE SUBANUNS OF SINDANGAN BAT 

By Emerson B. Chmsth 

Order No. 410. Paper, 121 pages, 1 
map, 29 plates. $1.25, postpaid. 

Sindansan Bay is situated on the north- 
ern ooast of Zamboanga Peninsula. The Su- 
banuns of this reaion were studied by Mr. 
Christie during two periods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanuns at 
work and at play; their industries, houses, 
altars, and Implements; and the people 
themselvr . . 



By Herbert S. Walker 

Order No. 412. Paper, 145 pagss, 10 
plates, 1 map, $1.25, postpaid. 

Considered from the viewpoint of prac- 
tical utility. Mr. Walker's Sugar Industry 
In the Island of Negros is one of the most 
Important papers published by the Bureau 
of Science. This volume Is a real contribu- 
tion to the subject; it is not a mere com- 
pilation, for the author was in tho field and 
understands the conditions of which he 
writes. 



THE HISTORY OF SULU 

By Najeeb M. Saleeby. 

Order No. 406. Paper, 275 pages, 4 
maps, 2 diagrams, $0.75, postpaid. 
In the preparation of his manuscript for 
The History of Sulu, Doctor Saleeby apent 
muoh time and effort In gaining access 
to documents in the possession of the Sultan 
of Sulu. This book is a history of the 
Moros In the Philippines from the earliest 
times to the American occupation. 



By Charles S. Banks 

Order No. 413. Paper, 53 pages, 20 
plates, $0.75, postpaid. 

In A Manual of Philippine Silk Culture 
are presented the results of several years' 
actual work with silk-producing larva to- 
gether -with a description of the new Philip- 



THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 



Vol. XII 



MAY, 1917 



No. 3 



DESTRUCTIVE DISTILLATION OF PHILIPPINE WOODS » 

By A. H. Wells 
(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

In the United States and other countries the destructive dis- 
tillation of waste wood has become a large and growing industry. 

In order to furnish some idea as to the possible introduction 
of such an industry into the Philippine Islands, certain classes of 
woods were selected and the relative yields of products wer£' de- 
termined on representative specimens. 

Table I. — Classification of commercial woods used for destructive 
distillation. 



Botanical name. 


Commercial name. Group."" 


Hardness. » 


Specific 
grav- 
ity, b 


Mois- 
ture.b 


• 

Rhizophora sp 

Bruguiera parviflora W. & A... 




4 
4 
4 
1 
1 
2 
3 

1 
4 
3 
4 


Very hard.. 

do 

do 


0.925 
1.053 
0.754 
0.933 
0.930 
0.716 
0.721 
0.492 
0.630 
0.458 
0.711 
0.687 


P. cent. 
8.1 
9.0 
7.7 
6.7 
7.2 
7.4 
10.0 
7.8 

8.6 
8.4 
9.2 






Intsia bijuga (Colebr.) 0. Ktze _ 


Ipil 






Shorea guiso (Blanco) Blume ... 

Dipterocarpus sp_. 

Shorea polyspermia ( Blanco) Merr 




do 


Apitong 


Medium hard 

do 

do 


Narra 


Pentacmecontorta (Vid.) M. & R. 
Anisoptera thurifera Blanco 


White lauan 


Soft 




Benguetpine 









" Group and hardness are taken from Annual Reps. P. I. Bur. For. (1906-1914). 
b Specific gravity and moisture determination were made on the specimens used. 

The percentage of charcoal made in the Philippines from wood 
other than that cut for such specific purpose is very small. 
There exists in the Philippines a vast amount of wood of a waste 
nature, the possible utilization of which might be a great benefit 
to certain industries. Throughout the Islands in the timber 



Received for publication January 



1917. 



112 The Philippine Journal of Science i»n 

regions the annual production of waste wood about certain mills 
has so increased that the question of its disposal is now a problem 
before the mill men. Another possible source of waste wood is 
that from the cutch, or tan bark, industry. During the fiscal 
year 1914 there were taken from the public forests of the Philip- 
pine Islands 2,793,295 kilograms of tan bark. 2 The highest 
percentage of this bark cdmes from the mangrove swamps. 
These swamps cover extensive areas in Mindanao, Palawan, and 
other islands in the southern parts of the Archipelago. If means 
for a profitable utilization of the wood from these trees could 
be devised, it seems probable that a stimulus would be given to 
a larger tan bark industry. 

The woods selected for distillation are representative of the 
timber trees that are found in greatest abundance and also of 
the trees furnishing tan barks. 

Bacauan, langarai, and api-api represent classes of woods 
which comprise the greater portion of the forest areas in the 
swampy regions of Mindanao. The bark stripped from trees 
of the bacauan family is especially of value as a source of tannins, 
while the wood itself is little used as lumber and offers a ready 
supply for distillation. The area of these mangrove swamps is 
approximately 207,200 hectares. 3 Apitong, tangili, guijo, yacal, 
white lauan, and palosapis belong to the Dipterocarpacese and 
are representative of some of the principal timber trees that are 
milled in the Philippines. The dipterocarp forests are estimated 
by the Bureau of Forestry as covering 70 per cent of the total 
area of the commercial forests of the Islands. Ipil and narra 
belong to the Leguminosae and are representative of that class of 
timber trees. Benguet pine, Pinus insularis Endl., is the com- 
mon pine, found principally in the Benguet region. It is milled 
particularly in northern parts of Luzon, where it extends over 
an approximate area of 518,000 hectares. 

The specimens of wood used for furnishing the analytical data 
were taken from several sources. 

The bacauan, the api-api, and the langarai were identified 
and cut in Mindanao swamps and were shipped to the Bureau 
of Science, where they were allowed to season for five years. 
The pine logs were obtained from the Benguet region through 
the courtesy of the Bureau of Forestry. The other specimens 
were identified and purchased from reliable lumber dealers in 
Manila, care being taken in selecting samples that seemed to 

"■Rept. P. I. Bur. For. for 191U (1915) and ibid, for 1915 (1916). 
'Cox, Alvin J., This Journal, Sec. A (1911), 7, 2. 



xii. a, 3 Wells: Destructive Distillation 113 

show a likeness in size, color, specific gravity, and apparent 
moisture content. 

A separate lot of each kind of woods was kiln-dried until the 
moisture content seemed uniform. The whole specimen was 
then placed in a tight container, and from this the daily charge 
was taken and distilled. From eight to thirteen charges were 
made from each specimen and specific gravity and moisture de- 
termination made on each charge- The averages for these de- 
terminations are found in Table I. The specific gravities were 
taken by the usual method. The measure of the moisture was 
not that of the total volatile substance below 108° C, which 
in many cases would contain oils, but a measure of the volatile 
part taken up by pure calcium oxide. The determination was 
made by passing the volatile matter through a tube of calcium 
oxide, afterward heating the tube for three hours at a temper- 
ature of 185° C. sufficiently high to volatilize all oils present and 
yet not to drive off water from the partially slaked lime. 

Destructive distillation was selected as the method for distill- 
ing the woods, the reason being that it seemed to give the 
highest yields of the commercial products — methyl alcohol, 
acetic acid, tars, and charcoal. Steam distillation might have 
shown interesting results regarding the resinous products 
obtainable, and this method may be applied later to certain 
classes of woods that are known to be high in resinous content. 
The destructive method is that in general use in the distillation 
of hard-wood waste, and for that reason also it is used here to 
furnish experimental data of comparative value. 

In determining a laboratory method for the distillation, it 
seemed advisable to have a temperature control. During 1913 
experiments were made in this laboratory by utilizing an oil- 
jacketed retort. In the use of the oil-jacketed retort a heavy 
hydrocarbon oil was heated and kept in circulation by the aid 
of an electrically controlled Kinney pump, which was connected 
to the retort so that it supplied cold oil from the storage tank 
and also accelerated the circulation of the hot oil about the 
retort. Coils of various sizes and forms were tried as heating 
surfaces for the oil. The most satisfactory surface seemed to 
be that of a piece of iron pipe 17 centimeters long by 13.5 centi- 
meters in diameter, capped at both ends, the caps being bored 
to accommodate a 1-inch circulation pipe. 

The heat was furnished by two Meker burners connected to 
the laboratory gas supply. The loss of heat by radiation was 
cut down by placing a heavy shell of asbestos cement over the 



114 The Philippine Journal of Science i9n 

retort. But this method gave neither the close temperature 
control nor other results desired. Also the cylinder oils sub- 
mitted for the bath gave unsatisfactory results under high tem- 
peratures; consequently the method was abandoned. 

In June, 1916, an improved experimental plant was set up 
and operated. The plant consisted of a cylindrical iron retort 
mounted at a suitable height and heated electrically in such a 
manner that the inside temperature could be closely controlled 
over a range in temperature of 500° C. The capacity of the 
retort was from 10 to 15 kilograms, varying with the specific 
gravity of the wood. The four heating coils were made of 
No. 18 "ni-chrome" wire arranged to give an approximate 
reading of 10 amperes for each coil. The wires were wound 
about the retort connected in parallels, insulated from the sides 
of the retort by a layer of asbestos board, and covered on the 
outside by a shell filled with a layer of asbestos powder 6 centi- 
meters thick. This outer layer left slight chance for loss of heat. 

The control of the temperature was effected by inserting re- 
sistance and by breaking the main circuit. Each coil in the 
resistance cut from the circuit gave a rise of 2 amperes current. 
This arrangement gave a very satisfactory temperature control. 

The condensation of the distillates and the scrubbing of the 
noncondensible gases occurred in a coil made of lead pipe 
2.5 centimeters (1 inch) in diameter fitted inside an ordinary 
barrel of strong material and connected on the outside to a 
small tower from which extended pieces of glass tubing 400 
centimeters long and of diminishing diameter. This tubing 
aided in condensing the lighter product and carried off the 
noncondensible gases. The tar trap for catching the heavy 
distillates was placed between the retort and the condenser. 
It consisted of a piece of iron pipe 10 centimeters (4 inches) 
in diameter, capped at both ends and fitted with a pet-cock at 
the bottom. Standardized thermometers were used in noting 
the temperature changes, which were taken at the center of 
the charge. Corrections for the emergent stem of the ther- 
mometer were made in accordance with the common formula 
0.00016 (T-t) N. 

In order to get comparative data, both fast and slow runs 
were made on the lots of woods. In operating the still, the 
weighed charge was placed into the retort and the current was 
applied. The fast or uncontrolled runs were made with 35 to 40 
amperes of a 110 volt direct current and occupied about nine 
hours before practical completion. No effort was made to 
control the heat except at the beginning of the exothermal reac- 



Wells: Destructive Distillation 



115 



tion, which began between 265° and 275° C. During this period 
the current was shut off until the reaction subsided in violence 
and was then immediately applied. The current was shut off 
to prevent a loss by overtaxing the condensing system. 

In the slow, or controlled, distillation the current was so ap- 
plied that the apparent moisture was first driven off. The 
moisture was considered driven off when a temperature was 
reached at which there was a decided pause in the distillation. 
This pause occurred at different temperatures for different classes 
of wood and varied between the temperatures 121° and 180° C. 
at the center of the retort. Later the passage of more current 
was allowed so as to afford the minimum temperature for dis- 
tillation. At the beginning of the exothermic distillation the 
main circuit was broken, and no current was allowed through 
until the rise in temperature ceased and distillation lessened. 
The rise in temperature during this period seldom exceeded 
20° C. After passing this period of exothermic reaction, the 
distillation dropped off rapidly and the application of current was 
gradually increased up to a temperature of about 330° C. and 
was increased rapidly to 550° C. In collecting the distillate, 
graduated cylinders were used. 

Samples of pyroligneous distillate from each distillation were 
analyzed for percentages of wood alcohol and acetic acid. Such 
analyses taken on fast runs were compared with analyses of 
products from controlled runs, and the results were tabulated 
in averages. 

Table II. — Showing the average percentages of wood alcohol and acetic 
acid obtained from a fast and from a slow distillation. 



Wood. 


Slow distillation. 


Fast distillation. 


Methyl 
alcohol. 


Acetic 
acid. 


Methyl 
alcohol. 


Acetic 
acid. 




0.91 

1.20 
1.55 

0.74 
1.61 
1.36 
0.82 
2.12 
1.71 
1.84 
1.01 


3.80 
2.92 
2.90 
4.70 
4.61 
3.83 
4.40 
4.37 
2.12 
5.16 
4.66 

6.31 


0.72 
1.10 
1.11 
1.31 
0.92 
0.62 
1.41 
1.24 

1.56 
1.42 
1.17 
1.00 


3.91 
2.85 

5.18 
4.25 
3.62 
4.42 
4.77 
2.51 
4.84 
4.27 
4.50 
6.16 








Yacal - 




Ipil 

















116 The Philippine Journal of Science i9n 

The percentages in Table II are calculated on the moisture- 
free sample and not on the sample as kiln-dried wood. The 
figures expressing wood alcohol also include slight percentages 
of acetone present. Klar's 4 method was used in the chemical 
analyses, together with those used by the United States Bureau 
of Chemistry 5 and the Philippine Bureau of Science. 

The results expressed in Table II show that a distillation pro- 
ceeding under controlled temperature condition gives in most 
cases an increase in yields of methyl alcohol ; that is, the average 
percentage of alcohol obtained by slow distillation is in all cases 
higher than that obtained by fast distillation. The mangrove 
woods bacauan, api-api, and langarai gave the highest yields in 
alcohol and acetic acid. Of the twelve woods distilled, those 
highest in resinous content gave the lowest percentages of these 
two products. The figures obtained for the yields of acetic acid 
do not show a relationship constant enough to warrant the state- 
ment that the controlled distillation gives higher percentages 
of acetic acid, although the greater number of distillations 
high in acetic acid are those carried out under the controlled 
distillation. 

In most cases the distillate began passing over at 110° to 
117° C. and continued at approximately the same rate up to 
265° to 275° C, when the exothermic action of the decomposition 
tended to accelerate the flow. The clear liquor passing off below 
180° considered as moisture gave no qualitative tests for the 
presence of alcohol, but did show an acid reaction toward litmus. 

Distillation under temperature control on certain hardwoods 
in the United States shows interesting results. Lowering the 
temperature of the reaction and decreasing the speed of the dis- 
tillation in the laboratory at the period of exothermal reaction in- 
creased the yield of methyl alcohol 45 per cent. The application 
of this method to a commercial retort indicated possible yields 
of alcohol 30 per cent higher than by the usual methods of firing 
and an increase of 15 per cent in yields of acetate of lime. The 
best results were obtained by first rapidly removing the moisture 
content of the charge and later decreasing the heat at a period 
just before the destructive distillation began. This method of 
temperature control gave promise of being entirely applicable in 
the commercial plant. 

* Klar, M. Technologie der Holzverkolung. Julius Springer, Berlin 
(1910). 

6 Circ. U. S. Dept. Agr., Bur. Chem. (1907), No. 63. 
"Palmer, R. C, Journ. Ind. & Eng. Chem. (1915), 7, 668. 



Wells: Destmctive Distillation 



117 



The yields from the destructive distillation of various Philip- 
pine woods have been tabulated and placed below. All of the 
calculations are expressed as percentage yields based on the 
moisture-free wood. The pyroligneous distillate represents the 
acid liquor free from tars. The percentage gas is taken by 
difference. 
Table III. — Percentage yields of various products from Philippine woods. 



Wood. 


3 . 


w 


i 

5 


3 

c 

1 

E 


C m 

gs 
II 


i 


• c 8 

s 


§ c3 

- 3 as 
< 


o 
O d 

■° 5 

3 

2 




39.8 
36.6 
35.5 
34.6 
39.9 
32.4 
39.4 

39.5 
40.8 
43.4 
41 .3 


6.2 
4.8 
3.6 

3.4 
2.6 
3.9 
5.7 
5.5 
2.7 
2.9 
3.4 
6.9 


5.6 
2.7 
3.2 
2.3 
2.1 
1.2 
2.0 
2.1 
3.6 
1.4 
1.3 
1.8 
3.2 


8 

i.i 

.9 

1.2 


15.1 
18.8 
23.4 
20.0 
20.3 
19.7 
19.1 
16.0 
15.1 
21.2 
22.9 
19.0 
16.2 


32.6 
33.9 

38.3 

36.6 
41.7 
36.8 
35.7 
35.2 
32.1 
32.4 


.91 
1.23 
1.20 
1.55 

.93 

.74 
1.61 
1.36 

.82 
2.12 
1.84 
1.71 
1.0 


2.92 
2.90 
4.70 

4.61 

4.40 
4.37 
2.12 
5.16 
4.95 
4.66 
6.31 


13 
8 
8 

8 

8 
8 
8 
8 

8 
5 






Guijo 


Yacal... 




Ipil ....'. 




Bengiiet pine 

Bacauan 









The pyroligneous distillate as it comes from the retort ranges 
in color from pale straw, given by palosapis, to a deep brownish 
red, given by narra. Titrations were made on ten fractions 
covering the periods of complete runs. The fraction showing the 
highest percentages of acid varied with the class of wood dis- 
tilled, but fell between the temperatures 285° and 325° C. Table 
III gives the yields expressed as 100 per cent acetic acid on dry 
weight of wood based on a slow or controlled distillation. 
Bacauan, langarai, and api-api give the highest percentage yields 
of methyl alcohol and acetic acid. The woods giving the lowest 
yields are woods known as soft woods and woods containing 
resin. 

Results obtained in the United States showed 7 yields of acetic 
acid for maple, beech, and birch (three hardwoods) as high as 
6.30, 6.28, and 6.96 per cent, respectively, the percentages being 
based on per-unit weight of the oven-dried wood distilled. It 
may be that distillation of other classes of Philippine hardwoods 
will show higher percentage yields than those obtained from 
the woods already examined. 



' Palmer, loc. cit. 



118 The Philippine Journal of Science im 

Determinations of acetone in the pyroligneous distillate were 
neglected except for qualitative tests showing its presence. 
Many of the distillates, especially those from the resinous woods, 
contained aldehyde and higher ketonic bodies and substances 
influencing the reduction of the iodide. For the determination 
of the alcohol the method suggested by Stritar and Zeidler s 
was used, but was afterward replaced by the modified methyl 
iodide method. With the hardwoods free from light oil the 
percentage could be taken from the specific gravity of the dis- 
tillate according to the methods of Klar. 9 



The tars were classified as settled tars, dissolved or "boiled" 
tars, and floating tars. 

The settled tar was the portion taken from the tar trap added 
to that which settled at the bottom of the pyroligneous liquor. 
Dissolved tars are those lighter portions dissolved in the pyro- 
ligneous liquor, but separated by boiling off. The floating tars 
were very light products found floating on the surface of the 
distillate and in the scrubbing tower. They gave tests for 
methyl alcohol, acetone methylated compounds, ketones, terpenes, 
and neutral hydrocarbon oils. The total tar included all three 
of the classes, and as such it was used as an analytical sample 
in the case of each wood distilled. 

The composition of oils originating from the destructive 
distillation of wood has been thoroughly studied by Fraps. 10 
Working on tar from hardwood, he arrived at the following 
results, which show the complex nature of a hardwood tar : 

(1) The wood oil examined is a mixture of a homologous series of dif- 
ferent classes of compounds. 

(2) It probably contains the series of aldehydes CnH2nO, corresponding 
to the series of saturated acids CnH^nCX', found in wood vinegar, which goes 
up to and includes caproic acid. The quantity of aldehydes is so small 
that only one could be positively identified, valeric aldehyde. Acetic alde- 
hyde and propionic aldehyde have been found in wood oil by other investi- 
gators, and these facts justify the first statement. It contains furfural in 
minute quantity. 

(3) It contains the series of ketones CnHanO, of which dimethylketone, 
methylethylketone, methylpropylketone, and probably methyl-H-butylketone 
and diethylketone have been detected. 

(4) It contains the ring ketone, cyclopentanone, and other ring ketones. 

8 Zeitschr. f. aval. Chem. (1904), 43, 381. 

9 Loc. cit. 

"Fraps, G. S., Am. Chem. Journ. (1901), 25, 26. 



Wells: Destructive Distillation 



119 



(5) It contains no mesityl oxide nor methylcyclopentanone. 

(6) It contains nitriles in minute quantity. 

(7) It contains the methyl esters of the saturated acids, CnH2nO;>, found 
in wood vinegar, including methyl acetate, methyl propionate, methyl 
n-butyrate, methyl ?i-valerate, very probably methyl caproate, and methyl 
heptoate and possibly methyl isobutyrate. The acid corresponding to methyl 
heptoate has not yet been found in wood vinegar. The percentage of 
ethereal salts decreases as their boiling-points rise, the lower-boiling ones 
being present in considerable quantity, the higher ones in very small 
quantity. One might conclude from this that the quantity of the corres- 
ponding acids in wood vinegar decrease with increase in boiling-point. 

(8) It contains the esters of unsaturated acids. 

(9) It contains methylfurane, and furanes which can be hydrolyzed by 
cold, strong hydrochloric acid to diketones. The furanes include dimethyl- 
furane, trimethylfurane, and higher furanes. 

(10) It contains no dimethylacetal, nor appreciable quantities of pyri- 
dines or alcohols. 

(11) It contains compounds which take up hydrochloric acid and bromine, 
and are therefore unsaturated. 

(12) It contains the aromatic hydrocarbons, toluene and xylene, and 
probably cumene and cymene. 

(13) It contains creosote. 

(14) It contains phenol ethers in minute quantity. 

(15) The higher boiling oils, freed from ketones, aldehydes, ethereal 
salts, and creosote can be separated into two parts by glacial acetic acid, the 
one soluble in that substance, the other insoluble. Both absorb bromine 
readily. 

Samples of tars from the Philippine woods were fractionated 
in a high distilling flask without the use of the Hempel column, 
and the results obtained are tabulated in average. 

Table IV. — Showing average percentage of light and heavy oils passing 
over in fractionation of the various tars. 

[Numbers give percentages.] 



Bacauan 

Langarai 

Api-api 

Ipil 

Yakal ._ 

Guijo.- 

Apitong 

Tanguili 

Narra 

White lauan. 

Palosapis 

Benguet pine 
Coconut shell 



Wood oil 
below 
100° C. 



Light 

oils, 100°- 

150° C. 



Heavy 

oils, 150°- 
250° C. 



28.0 

21 

64.0 



Distillate 
above 

250° C. 



120 The Philippine Journal of Science mi 

The samples taken for fractionation were freed from water 
by settling and by repeated use of separatory funnels. Yet on 
fractionation a high percentage of water was found dissolved in 
certain of the tars. In the cases of palosapis, white lauan, pine, 
and guijo the percentage of water ranged from 20 to 34, while 
in the others it seldom was above 10 per cent and was even as 
low as 2.6, as in api-api. 

The fractions passing over below 100° C. have a specific 
gravity of 0.958-0.998; are colorless, quickly turning yellow to 
brown; are very inflammable; and have a strong, characteristic 
odor. They gave reactions for aldehydes and ketones by the 
acid sulphite method. Attempts to form terpene compounds 
from the fractions from the resinous woods gave slight results. 
A few crystals of terpene hydrate were obtained. 

The fraction passing over between 150° and 250° C. and con- 
taining the creosote fraction ranged from pale yellow to brown. 
All specimens quickly darkened. The specific gravity varies from 
1.010 to 1.028. The green and the blue oils began passing over 
at 230° C, accompanied by a small amount of water, indicating 
decomposition of some of the compounds present. This decom- 
position and liberation of water was greater with the tars from 
the apitong and other resinous woods than from the bacauan and 
less resinous ones. The green anthracene oil preceded the blue 
coming over in distillation. The fractions from each of the 
tars showed marked reaction with caustic soda and gave rapid 
reaction for esters. Two cubic centimeters of the oil added 
to 100 cubic centimeters of a 20 per cent solution of sodium 
hydroxide set to a heavy, flocculent white precipitate. About 
70 per cent of this compound decomposes upon washing free 
of alkali or treatment with solvents. The percentages of oil 
unacted upon by the sodium hydroxide seemed very low. No 
further work was done on this portion of the tars. It is pro- 
posed at a later date to fractionate larger quantities of this 
fraction, getting the exact creosote fraction and the percentage of 
phenolic compounds present. The fractions passing over above 
250° C. are semisolid and usually red. They contain the highest 
fractions of the wood tar, together with certain high boiling 
fractions of the wood resins and their decomposition compounds. 
The residue after fractionation is a vitreous black pitch, having 
a conchoidal fracture and rather high melting point. It is 
slightly soluble in alcohol, chloroform, and ether and is moder- 
ately soluble in carbon disulphide, acetone, benzol, and xylene. 



Wells: Destructive Distillation 



121 



The charcoal taken from the retort is steel gray to black. It 
has a metallic luster and a conchoidal fracture. The charcoal 
from all the specimens, excepting tanguili, lauan, pine, and palo- 
sapis, is heavy and has a decided metallic ring when dropped 
upon a hard surface. The yields are from 32.1 to 41.7 per cent, 
calculated on the moisture-free sample. All of the specimens are 
hard to ignite and burn without flame, smoke, or odor of tarry 
matter. This class of charcoal would fill all the requirements of 
a charcoal for domestic purposes, while some of the woods would 
furnish an excellent charcoal for the iron industry in the 
Philippines. 

Table V shows the results of an analysis of bacauan charcoal 
that came from a charge heated to 550° C. The sample was 
taken from an old lot of bacauan charcoal that had stood in 
the laboratory for a year. 

Table V. — Analysis of bacauan charcoal. 



CHARCOAL. 



Moisture (110° C.) 
Volatile combustible matter 
Fixed carbon 
Ash 

Total 
Available calories 
Gross calories 



Per cent. 

5.71 

23.79 

67.21 

3.29 

100.00 
6799 
7242 



Silica (SiO,) 

Iron and aluminum oxides (R2O3) 

Calcium oxide (CaO) 

Magnesium oxide (MgO) 

Potassium oxide (K 2 0) 

Sodium oxide (Na 2 0) 

Phosphoric anhydride (P 2 Os) 

Sulphuric anhydride (SOs) 

Chlorine (CI) 

Carbon dioxide (undetermined) 



Analyzed by A. S. Argiielles. 



Per cent. 

2.98 

2.95 
61.81 
2.72 
2.43 
8.93 
1.21 
2.29 
0.12 
14.56 

100.00 



The volatile combustible matter seems somewhat high. The 
calorific values are excellent and probably higher than any 
value that could be obtained on charcoals made by the Filipino 
process of burning. The analysis of the ash indicates that 
bacauan charcoal might be unsuitable for the manufacture of 



122 The Philippine Journal of Science im 

pig iron for use in the foundries in the Philippines, due to the 
high percentage of sulphur which it contains. 

Cox " has shown that there are many woods in the Islands 
from which the charcoal obtained is entirely free from sulphur 
and is suitable for use in the smelter. Certain features shown in 
the analysis of the ash, such as the high percentages of calcium 
and sodium oxides, may be possibly explained by the fact that 
the mangroves are distinctly salt-water species occurring in 
tidelands. The high calorific value places bacauan charcoal 
as a high-grade fuel for all domestic purposes, and for such 
purposes there is a great demand for charcoal in the Philippines. 

Coconut-shell charcoal possesses remarkable absorptive prop- 
erties. Such charcoal was used by Wright and Smith 12 for filling 
the absorption tubes used in their quantitative determinations 
of radium emanation. 

The matter of first cost and operation expenditures is one of 
individual calculation, depending upon certain local conditions. 13 

The production costs, costs of installation, and production 
values received in the United States in the distillation of hard- 
woods have been discussed in some detail. 1 * 

French 15 finds the production cost to be approximately 17.70 
pesos and the production value of crude products 19.82 pesos, 
giving a profit of 2.12 pesos per cord. He places the cost of 
installation, eliminating the cost of wood supply, at 4,000 pesos 
per cord per day production. These figures are for a plant 
producing crude products from hardwoods. 

In determining whether a plant may prove profitable or not, 
many factors must be taken into consideration. 

Through the courtesy of a manufacturer of wood-distillation 
plants 16 this Bureau has been furnished with a few of the 
leading questions that should be considered as important factors 
in plans for a wood-distillation industry. 

1. How much wood, and of what quality, is to be worked per year, and 
in how many working days? What is the price of sufficiently split wood 
at the place where the factory is situated? 

"Cox, A. J., Philippine firewood, This Journal, Sec. A (1911), 6, 10. 
"Wright, J. R., and Smith, O. F., This Journal, Sec. A (1914), 9, 54. 
"Klar, loc. cit. 

"French, E. H., Journ. Ind. & Eng. Chem. (1915), 7, 55. 
" One peso Philippine currency equals 100 centavos, equals 50 cents 
United States currency. 

" Mr. F. H. Meyer, Hannover-Hainholz. 



xii. a, 3 Wells: Destructive Distillation 123 

2. What is the weight per cubic meter, or what is the form of the wood 
used? How long does it lie before it is put in work? 

3. What products are principally intended to be manufactured: 
Charcoal. Anhydrous tar. 

Charcoal bricks. Tar oils and creosote. 

Brown acetate of lime. Crude wood spirit of turpentine. 

Gray acetate of lime. White deodoriferous pine oil. 

Crude wood spirit. Pine tar oil. 

Refined wood spirit. Crystallized acetate of sodium. 

Pure methyl-columbia spirit. Acetate of sodium, free of water. 

Acetic acid in all kinds of quality. Formaldehyde. 

Acetic acid for vinegar. Paraformaldehyde. 

Acetone. 

4. Is there a market for charcoal, and at what price? 

5. Is there a market for tar, and at what price? 

6. What kind of fuel may be had and at what price? 

7. Is cooling water to be had (state whether fresh or salt water), and 
what is the mean temperature? 

8. Is running water in the neighborhood? 

9. Describe the local conditions (distance to nearest railroad station 
or steamship landing place, harbor, etc., as well as character and condition 
of roads, etc.). 

10. Is there any junction for railway or ship? 

11. What is the freight for the products as far as the next place of 
consumption? 

12. What is the freight for apparatus and machinery from manufacturer 
to place of factory? 

13. Are there any buildings to be utilized? 

14. Are there any repair shops and how large are they? 

15. Are there gratuitous plots of land existing? 

16. What are the prices for building materials, such as brick, wood, 
mortar, etc.? 

17. What is the price for finished buildings above the ground: in brick, 
in framework, in sheet-iron roofs? 

18. What is the price for high-furnace-masonry, of brick stone, or 
chamotte? 

19. What is the price for iron-fittings for furnaces per hundred pounds? 

It may be readily seen that many of these questions might 
be answered favorably to the installation of a plant in the 
Philippines, while at the present time others make it seem 
impossible as a commercial success. 

From a study of the statistics of the Islands it seems probable 
that a market for the products of a wood-distillation industry 
could be found in the Philippines and near-by ports. Wood 
alcohol is used commercially as a denaturant and as a solvent for 
fats, oils, and resins. It is also used in the manufacture of 
aniline colors, smokeless powder, and formaldehyde and in 



124 The Philippine Journal of Science 1917 

various other chemical industries. Acetic acid is used in the 
manufacture of compounds used in the dye and paint industries. 
Acetone is also made from the acetates formed in the process. 
Acetone is used commercially as a solvent for resins and other 
compounds; it is also used in the nitro cellulose industry and in 
the preparation of many pharmaceutical compounds. The tars 
obtained by destructive distillation can be treated and certain 
fractions used in turpentine substitutes, cheap varnishes, lubri- 
cating oils and greases, inks, leather, soap and cement industries, 
and for impregnating timbers and ropes. 

The crude tars from the Philippine dipterocarps, being high 
in percentages of resins and oils, should prove very satisfactory 
as an impregnating or coating substance for ropes and calking 
used in ship building. Also, as many are high in creosote frac- 
tion, their use as a disinfectant or spray for sanitary purposes 
would furnish a reasonably cheap and efficient substitute for the 
more expensive mixtures now employed throughout the Islands. 
The pitch can be used for coating purposes. 

Certain conditions are essential for the successful operation 
of a distillation plant. The first essential is a supply of the raw 
material of the proper quality and in large quantities that are 
easily accessible. This not being obtainable, the plant should 
be one accommodated only to the needs of the mill for clearing 
away the accumulation of waste wood and sawdust, which may 
hinder proper operation of the mill. However, in this case it 
must be observed that the use of small retorts of a capacity of 
0.25 to 1 cord would require a disproportionately greater ex- 
penditure for wages, for fuel, and for repairs than would the 
larger retort systems. Likewise the percentage yields are 
greatly reduced, due to the overheating in the small retort. 

A plant erected for supplying products to the Philippine 
market would probably manufacture charcoal, gray acetate, 
refined wood spirit, tar oils, and creosote. All of these prod- 
ucts have a market and could be used in Philippine arts and 
trades with the possible exception of the gray acetate, for which 
use might be made by its conversion into acetic acid or acetone. 

CONCLUSIONS 

Specimens from twelve kinds of wood have been submitted to 
destructive distillation. 

A furnace, heated electrically, has been devised to furnish 
controlled as well as noncontrolled distillations. 



xii, a, 3 Wells: Destructive Distillation 125 

The yields of methyl alcohol, acetic acid (100 per cent), 
charcoal, and tars have been tabulated to show average perr 
centages of products obtained by laboratory practice. 

Controlled distillation has shown results higher than those 
of the noncontrolled distillation in average percentage yields 
of methyl alcohol and acetic acid. These results are analogous 
to those found by Palmer. 17 

A short discussion of costs and operating expenses is added. 

17 Loc. cit. 



THE POSSIBLE MAXIMUM VITAMINE CONTENT OF SOME 
PHILIPPINE VEGETABLES 1 

By Harvey C. Brill and Cecilio Alincastre 
(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

Recent investigations have proved that a diet composed of the 
proper proportions of fat, protein, and carbohydrates is not 
necessarily an adequate diet for the maintenance of health and 
growth. Other ingredients must be present; for example, in- 
organic salts must be present in sufficient quantities. 2 

The absence of certain either-soluble 3 substances is prejudicial 
to health. This substance has been named "fat soluble A" by 
McCollum, 4 because of its presence in certain natural fats. 

Besides the above-required constituents, another substance, 
known under the various names "vitamine," 5 "oryzanin," G and 
"water-soluble B," T is necessary. 

No reliable quantitative methods for the determination of this 
vitamine constituent have been perfected. In the processes used 
for its isolation, much of it is undoubtedly broken up into simpler 
substances, 8 or the antineuritic principle may be due to the 
presence of a number of bodies. The latter assumption is con- 
siderably strengthened by the results obtained by the use of 
various substances 9 in the treatment of beriberi. A further 

1 Received for publication April, 1917. 

2 Osborne, T. B., and Mendel, L. B., Feeding experiments with isolated 
food-substances, Publ. Carnegie Inst. Washington (1911), No. 156, Parts 
I and II; and Science (1911), 34, 722. McCollum, E. V., and Davis, 
M., Journ. Biol. Chem. (1913), 14, 40. Hart, E. B., and McCollum, E. V., 
ibid. (1914), 19, 373. 

3 McCollum, E. V., and Davis, M., ibid. (1913), 15, 167. Osborne, T. B., 
and Mendel, L. B., ibid. (1913), 15, 332. 

4 McCollum, E. V., Simmonds, N., and Pitz, W., ibid. (1916), 28, 153. 

5 Funk, C, Journ. Physiol. (1912), 43, 395. 

"Suzuki, U., Shimamura, T., and Odake, S., Biochem. Zeitschr. (1912), 
43, 89. 

'McCollum, E. V., and Kennedy, C, Journ. Biol. Chem. (1916), 24, 491. 

5 For a brief bibliography of methods, see Williams, R. R., This Journal, 
Sec. A (1916), 11, 49. 

8 Williams, loc. cit. and Journ. Biol. Chem. (1916), 25, 437. Williams, 
R. R., and Saleeby, N. M., This Journal, Sec. B (1915), 10, 99. Williams, 
R. R., and Seidell, Atherton, Journ. Biol. Chem. (1916), 26, 431. 

149517 2 127 



128 The Philippine Journal of Science iw 

confirmation of the probability of the existence of a number of 
antineuritic substances in rice polishings, yeast, wheat, and 
bran, which are antineuritic in character, is the recent discovery 
by Williams 10 of the peculiar action toward polyneuritic pigeons 
of the hydroxypyridines, when they undergo dynamic isomerism, 
and of the similar action of the vitamines isolated from autolyzed 
yeast. 11 

If the vitamine content of foods does consist of a number of 
distinct chemical compounds, the difficulty of quantitatively de- 
termining it is readily seen. However, some attempts have been 
made to make such an estimate. Phosphotungstic acid in alka- 
line solution gives a blue color reaction with the antineuritic 
substances. 12 Phosphomolybdic acid gives a similar color reac- 
tion. 13 The intensity of the color is dependent on the concentra- 
tion of the vitamine. The handicaps attached to these methods 
are that no standard exists that can be used for comparison and 
that other compounds give similar color reactions. 

The quantitative isolation is unsuccessful because of the pres- 
ence of nitrogen compounds, which either accompany the vita- 
mine or are the result of its decomposition. Vedder and 
Williams 14 report that three successive extractions of rice polish- 
ing with alcohol did not extract all the antineuritic properties of 
the polishings. 

Funk 15 has made an attempt to determine the vitamine con- 
tent of milk in the following manner: The milk was distilled in 
a partial vacuum at 30° C. It was then powdered and dried to 
constant weight. The powder was shaken with alcohol for two 
hours and filtered, and the filtrate was evaporated to dryness. 
The residue was extracted with water, acidified with sulphuric 
acid, and treated with phosphotungstic acid. The nitrogen of 
the phosphotungstic acid precipitate was determined by the 
Kjeldahl method. By this method Funk estimates the vitamine 
content of milk to range from 1 to 3 centigrams per liter or 
from 0.001 to 0.003 per cent. 

"Williams, loc. cit. 

11 Williams and Seidell, loc. cit. 

12 Drummond, J. C, and Funk, C, Biochem. Journ. (1914), 8, 598. Folin, 
O., and Macallum, A. B., Journ. Biol. Chem. (1912), 11, 265; (1912), 13, 
363. 

13 Folin, O., and Denis, W., ibid. (1912), 12, 239. Funk, C, and Macallum, 
A. B., Biochem. Journ. (1913), 7, 365. 

M Vedder, E. B., and Williams, R. R., This Journal, Sec. B (1913), 8, 175. 
"Funk, C, Biochem Journ. (1913), 7, 211. 



xii, a, 3 Brill and Alinc'astre: Philippine Vegetables 129 

In a subsequent communication Drummond and Funk 1U report 
that all the nitrogen of vitamine is not obtained by means of 
the Kjeldahl method ; so the results recorded above for milk are 
probably low. Compounds containing the pyridine ring possess 
this same property of resisting digestion with sulphuric acid for 
the determination of nitrogen. 17 Approximately three fourths 
of the nitrogen of such bodies were obtained by this method. 

A method based upon this property of pyridine derivatives has 
been employed by the authors for the determination of the pos- 
sible vitamine content of some vegetables grown in the Philippine 
Islands. 

DETERMINATION OF THE VITAMINE CONTENT OF VEGETABLES 

A portion of the fresh vegetable was dried at a low temperature 
and very finely ground, and 100 grams of the ground foodstuff 
was exhaustively extracted with methyl alcohol. The alcohol 
was evaporated at low temperature, and the residue was taken 
up in water and filtered. The filtrate was acidified with suf- 
ficient sulphuric acid to make a concentration of 5 per cent. 
Phosphotungstic acid was added to this solution to precipitate 
the antineuritic substance. After standing for from four to 
eight hours, the precipitate was filtered off, washed with 5 per 
cent sulphuric acid and finally with alcohol, and dried in a 
desiccator over sulphuric acid. The nitrogen was determined 
in this residue by both the Kjeldahl and Dumas methods. 

Weighings of the precipitates and nitrogen determinations 
were made to serve as checks. These determinations are in- 
cluded in Table I. The calculations were made on the vege- 
tables 18 as they were prepared for the table. 

The values for the vitamine content of Philippine vegetables 
found by this method are usually higher than the results found 
by Funk 19 for milk. However, they are presented here because 
they have comparative value and for the purpose of stimulating 
further investigation in quantitative determination of these 
constituents. 

"Drummond, J. C, and Funk, C, Biochem. Journ. (1914), 8, 598. 

"Brill, H. C, and Agcaoili, F., This Journal, Sec. A (1917), 12, Kjeldahl 
process. 

18 For description of Philippine vegetables, see Agcaoili, Francisco, ibid., 
Sec. A (1916), 11, 91. 

"Funk, C, loc. cit. 



130 



The Philippine Journal of Science 



;?i 






§ 8 



8 8 

d © 



1 S I 1 1 1 

o © © o o o 
o d d o d d 



gooooooooooooo iOOOO 



ooooooooooo 



o o o o o 



gS358SS 



:!Islsl3 s 



f-ais 



S £ S3 B 

rt rJ d rH 












8 8 8 5 S S @ 



rioOrHO^OOOOOOrtrtrtrtOO 



Ola !**• j 



1«e:??§SS§ 



S : § 8 



O O r-H O (N © 



rt ..HO 









!£gggS§E553 

, d d d d d d cm' d d d d d d 



Oir-tco©U3INCOrr 



i i i i ■ i i k, g> > 

i i I J I i -O ^ | 

| g | j j I -S § £? | ^ ^ 2 



O W K M 



■J * S 2 2 c a ? S J! a a • - 

S 1 .s S | | s | g i ^ t g s 
ojw(2mouwS5mm(5h 



Brill and Alincastre: Philippine Vegetables 



131 



it 

I 1 " 



S 8 

d © 



to " 
. c 

« ^ 

. o 

PS g 

g "S 



g -o 



I 1 

.-! d 



- « "" g 

5 ■= * 3 
~ ° g I 



j 2 



SI,*** 



.■a s: " £ o t; 

S £ « -o c 5 



S| 



5 2 E 

ell 



G ^ m . a 



132 The Philippine Journal of Science 

SUMMARY 

Methods for the quantitative estimation of the vitamine con- 
tent of foods are discussed. 

A method based on the different values for nitrogen in py- 
ridine ring bodies obtained by the use of the Kjeldahl and the 
Dumas methods is presented. A table with some results for 
Philippine vegetables is given. 



THE EFFECT OF CALCIUM SULPHATE ON CEMENT * 

By J. C. Witt and F. D. Reyes 

{From the Laboratory of General, Inorganic, and Physical Chemistry, 
Bureau of Science, Manila) 

Calcium sulphate in the form of either gypsum or plaster of 
Paris is almost universally used to control the setting of cement. 
The practice has led to extensive research in the attempt to 
determine the reaction between the two substances and the 
amount of calcium sulphate that can be added without causing 
harmful effects. As a result of these investigations some points 
are fairly well established. A number of writers 2 agree that 
the retardation of set caused by the addition of calcium sulphate 
is proportional to the amount added, only within certain limits ; 
that is, the setting time cannot be increased without limit by 
adding more and more of this substance. On the contrary, after 
a certain point is reached, further additions cause an acceleration 
of the set. 

Many believe that the permissible amount of calcium sulphate 
added to a cement should depend to some extent on the condi- 
tions to which the concrete made from it is to be exposed. 
Several countries, among which are France, 3 Japan, 4 and 
Argentina, 5 specify that a cement, intended for construction 
exposed to sea water, shall have a lower sulphuric anhydride 
(S0 3 ) content than one to be used under ordinary conditions. 

The German Portland Cement Manufacturers' Association ° 
recommended that "a uniform permissible maximum limit of 
S0 3 , namely 2.5 per cent, be generally adopted in Specifications 
for Portland Cement, whatever may be the purpose for which 

1 Received for publication November 13, 1916. 

2 Carpenter, R. C, Eng. Digest (1908), 3, 385. Rohland, P., Stahl u. 
Eisen (1908), 28, 1815. Reibling, W. C., and Reyes, F. D., This Journal, 
Sec. A (1911), 6, 225. 

"Ciment (1912), 17, 213. 

*Mitt. Zentralstell. Ford. Deut. Port. Zem. Ind. (1912), 1, 167. 

3 Ibid. (1912), 1, 305. 

6 Report of the International Association for testing materials (1912), 
Sec. 2, article XVIL, 1. (Translated by G. Salter.) 

133 



134 The Philippine Journal of Science im 

the cement is intended." This recommendation was made* on 
the basis of results obtained with only two cements : 

The materials used consisted of a Portland cement. (S), containing 1.19% 
of SOs and employed for all purposes, marine structures included, and 
another Portland cement (B), which contained only a very small proportion 
of S0 3 , namely 0.57%. 

In both cases the raising of the SCvcontent to 2.5% by the addition of 
gypsum, increased the tensile strength, both in fresh and sea water. 

In the summary of the paper it stated that — 

From these results it follows indubitably that the presence of up to 2.5% 
of SO3 in Portland cement, produces no injurious effects of any kind, 
whether in sea water or fresh water. 

No matter how conclusive these results appear, or how care- 
fully the work was carried on, it is impossible to settle such an 
important question by the behavior of only two cements, par- 
ticularly in the face of contradictory evidence obtained by other 
investigators. 

It is generally conceded that an excessive amount of sulphuric 
anhydride in a cement is harmful. All specifications mention 
an upper limit, though not all agree what this limit shall be. 7 
Meade 8 says : 

Although the presence of calcium sulphate in small quantities is bene- 
ficial to cement, there is no doubt that a quantity exceeding 4 or 5 per cent 
is injurious. 

According to Kiihl, 9 the sulphuric anhydride content of a 
cement should not exceed 2 per cent. Rigby 10 states : 

In making some experiments with cement by adding plaster of Paris by 
mixing the two materials in different proportions, I found that if I exceeded 
41 per cent of the latter the briquettes subsequently made from the mixture 
either broke up after being placed in water for a time, or, if they retained 
their shape, they were very much cracked and gave a very poor test. 

Bates lx found that while higher sulphuric anhydride content 

' Though the specifications of various nations for the upper limit of 
sulphuric anhydride are often taken as directly comparable, this is not the 
case. For instance, in the British specifications, 2.75 per cent sulphuric 
anhydride does not refer to the sulphur present as sulphate only, but means 
the total sulphur calculated to SO*. (See British Standard Specifications 
for Portland Cement. Revised, March, 1915.) 

s Meade, R. K., Portland Cement. Chemical Publishing Co., Easton, Pa. 
(1906), 31. 

9 Kiihl, H., Mitt. Zentralstell. Ford. Dent. Port. Zem. Ind. (1913), 2, 108. 

10 Rigby, J. S., Journ. Soc. Chem. Ind. (1890), 9, 254. 

"Bates, P. H., Proc. Am. Soc. Test. Materials (1915), 15, II, 126. 



xii. a, 3 Witt and Reyes: Calcium Sulphate on Cement 135 

(up to 2.5 per cent) in some cases increased the strength of 
neat briquettes, it caused considerable expansion. In attempt- 
ing to compare some of the papers on the subject, considerable 
confusion arises from the terminology of various writers. 
Some of them refer to the percentage of gypsum added to cement 
and others to the percentage of plaster of Paris ; as a rule, they 
do not state the percentage of sulphuric anhydride. Others 
mention simply calcium sulphate, and the reader has no means 
of knowing in which form it was added. Although the effects 
of the two substances are similar, 12 there is a considerable dif- 
ference in the amount of sulphuric anhydride in the commercial 
products. Different cements require varying amounts of calcium 
sulphate for the purpose of controlling the set. However, there 
must be some limit specified for the addition of material sub- 
sequent to calcination to avoid adulteration by unscrupulous 
manufacturers. The specifications state the maximum amount 
that can be allowed with safety. It remains to investigate 
the region between this amount and that definitely known to 
be injurious. It seems likely that the specifications of the 
countries that make allowance for the conditions to which con- 
crete is to be exposed show an advance in the right direction 
and that in addition to this there should be some relation be- 
tween the composition of the cement and the sulphuric an- 
hydride allowed. Until the subject is better known, safety 
demands that upper limits should be kept well below the amounts 
known to be harmful. 

The present work was undertaken to determine the effect of 
various amounts of calcium sulphate on several cements on the 
local market and to study in some detail the behavior of a 
cement made from raw materials available to the Bureau of 
Science for investigation. The finished cements investigated are 
from five different factories and are herein designated as A, B, 
C, D, and E. Each contained a certain amount of calcium 
sulphate placed there by the manufacturer. Various amounts 
of plaster of Paris were added to each, giving increasing amounts 
of sulphuric anhydride up to about 10 per cent. Investigation 
of the other cement, designated by F, started with the clinker, 
which contained only a trace of sulphuric anhydride. The set- 
ting time on each sample was determined as a preliminary test. 
These were used as a guide to determine the percentages of 
sulphuric anhydride best suited to extended tests. Tables I and 

"Meade, op. cit., 307. 



136 



The Philippine Journal of Science 



II show the analyses and the physical tests made on the cements 
as received. 

Table I. — Chemical analyses of cements as received." 

[Numbers give percentages.] 





Cement— 


A. 


B. 


C. 


D. 


E. 


F. 




2.50 
0.30 

21.10 
8.76 
1.42 

63.10 
1.26 
0.67 
0.93 


2.80 
0.60 

20.00 
8.86 
1.34 

63.00 
1.34 
0.86 
1.16 


1.50 
0.30 

22.50 
7.58 
1.12 

63.00 
1.74 
1.34 
0.87 


2.40 
0.80 

18.80 
9.18 
1.12 

64.10 
1.12 
1.34 
1.16 


2.00 
0.45 

18.95 
9.52 
1.38 

63.80 
1.80 
1.24 
0.87 


22.32 
8.21 
3.64 

63.15 
1.62 

trace 
1.09 




Silica (Si02) 












.Potassium and sodium oxides (K2O, Na20) 



Most of these analyses were made by Francisco Pefia, inorganic chemist, Bureau of Science. 

Table II. — Physical tests of cements as received. 



Per cent. 
87.6 



Per cent. 
99.0 



9 6 

6 40 

7 16 



Tensile strength b — kilograms per square centimeter. 



Neat cement. 



1 day. 7 days. 28 days. 60 days. 



1 day. 7 days. 28 days. 60 days. 



24. S 
23.6 
23.2 
24.5 
27.3 



41.2 
44.5 

44.2 



47.8 
50. 7 
51.3 



*.:-> 



30.9 
29.1 



Brand F, without any addition of calcium sulphate, was quick-setting ; consequently no 
test was made on the original material. 

b Each result is the average of six briquettes. United States Government specifications 
were followed. 



xii. a, 3 Witt and Reyes: Calcium Sulphate on Cement 137 

Table II. — Physical tests of cements as received — Continued. 



Brand." 


Ten9ile strength— pounds per square inch. 




Neat cement. 


Mortar 1 : 3. 


1 day. 


7 days. 


28 days. 


60 days. 


1 day. 


7 days. 


28 days. 


60 days. 




353 
337 
330 
348 
388 


695 
617 
586 
633 
628 


685 
651 
679 
722 
731 


673 
649 
644 
677 
720 


100 
100 
133 
124 
154 


238 
248 
282 
294 
331 


333 
308 
380 
409 
407 


360 
825 
428 
441 
414 


B 


C... 

D 

E 





* Brand F, without any addition of calcium sulphate, was quick-setting ; consequently no 
test was made on the original material. 

The plaster of Paris contained 55.88 per cent sulphuric an- 
hydride. Each cement was analyzed, and the approximate 
amount of plaster necessary to give a certain percentage of 
sulphuric anhydride was calculated. The two substances were 
then placed in a ball mill and thoroughly mixed. In the case 
of sample F the clinker was crushed to pass a 20-mesh sieve, 
mixed with the plaster, and ground to the fineness indicated in. 
Table III. This table shows the effect of sulphuric anhydride 
content on the setting time and on the tensile strength for va- 
rious periods up to ninety days. 

The results in Table III are not entirely uniform; however, 
as the results were obtained from more than 1,000 briquettes, 
general conclusions can be based on averages as follows : 

Setting time. — In conformity with the results, already men- 
tioned, of other investigations with calcium sulphate, most of 
the cements show a maximum retardation with 2 per cent or 
less of sulphuric anhydride and a shorter setting time with a 
higher or lower percentage. 13 In general, the initial-set curve 
obtained by plotting the percentage of sulphuric anhydride 
against the time is parallel to the final-set curve for the same 
cement. 



" Calcium chloride, sodium sulphide, and several other electrolytes have 
somewhat the same effect. [See Carpenter, R. C, Eng. Rec. (1904), 50, 
769. Witt, J. C, This Journal, Sec. A (1916), 11, 273.] 



138 



The Philipjrine Journal of Science 



S8S5S?£ 



SS^SS 





COOBWOhOiflflMX ■ <£> C- ii <£> ■ .-H CO CO *<* 

■^eo^ciciococoi— «•** i co co «-i ■** itDeoN^ 

>Ct-«0(OOlO<0(OIC i (O C C- (D ' tO C- <0 O 


si 


OOOi-lCOCOCJOCO^NO>OOCT>i-HNC300NCO 
CO^J'CONOi'-<OOr-<COOt-LO(M(r)CiWCMr-C-iO 






13 


rHcoomot-Nuaoo^o^fccvOi-HoocncotON 



i 



m 


OOTlOtOOOrHOO-aiC-OC-C-NOOt-OOOOOO 


•a 




"1 


15.9 
15.2 
13.0 
9.9 

15.0 
14.8 
18.8 
9.5 
10.0 
19.8 
17.0 
15.0 
9.0 
7.9 
20.7 
18.7 
18.3 
11.3 
10.4 



ss 4 



SSSSS :SS3« 



3 S S ss 



M*Tj"Tj«'*J'COU3'>J'^l--<3' 






o « » o 



53 w t- 



S 2 B 

SQ « * 






xii. a, 3 Witt and Reyes: Calcium Sulphate on Cement 139 





i 3 


ej 


3 


1 


c:> 


1 


II 




iii 


5 


| 


to 


■3 


1 s 


CO 




















3 


a s 1 


2 


s 


8 


a 


i 


S 


§ 


1 

er 
? 




§g 


1 


3 


I 


i 


s 1 


I 


> 




















■M 


" s s 


- 


s 


8 


j<5 


a 


on 


CO 


& 






































































1 s 1 


1 5 


S S 


£ 


3 




CJ 


















.c 




















1 1 § 


1 


1 


3 


V 


CO 


1 


s 


1 




o o 


o 


r- 


CO 


eg 


CO 


US 


N 


c 




00 r4 








a 








5 






































H 


to a m 


to 


cc 


N 


00 


<N 


n 


CO 


« 


g^ 


~f, 


s 


a 


s 




S3 


- 


s 


N •* rH 


CO 


to 


CM 


t- 


_ 


to 


en 


X 


















si 


CI i-l IN 


'"' 




' H 




M 


,H 




a 




en C- 


us 


H 


CO 


CO 


rl 


c- 


o 






3 s 


s§ 


?P 


" 


S !§ 5 S 




Tf <D •* 


N 


01 


H 


o 


O 


IO 


CO 




3 $ 3 


s 


s 


" 


5 


IO 


i§ 


S 


. 




















ea co to 


CO 
































3 




































N u n 


c- 


O 


o 


CO 


CO 


CB 


en 


« 


ass? 


■s 


S3 


a 


» 


s 


3 


53 


■a 




















N CXI tO 






o 




o 








§8 8 5 


'2 


5 


CO 


■w 


f: 


S 


8 


73 

C 


o o o 


to 


o 


O 


*» 


00 


CM 


•* 


v 


s """ s " 


88 


s 


en 


s 


sss 


p 


£ 


O 


p 


s 


O 


a 


1 


to to to 


to 


•o 


00 


e- 


CO 


CO 


IO 


c 

O 


Sggg 


3 


© 


- 


s 


8 


a 


■8 


•* ■* CO 


CO 


CO 


■* 


•* 


CO 


N 


N 


£ 




















S § o 


cb 


3 


p 


p 


p 


§ 


8 


£ 


CO CO CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


?! 


Tf tO CO 


m 


25 


... 





m 


,_, 




SB 


















H N M 


■* 


en 


rt 


N 


CO 


IO 


en 


« 




































c 






















































w 


















■ 


H 










fa 


























c 
































































m 









140 The Philippine Journal of Science 1917 

Tensile strength. — When 1.50 to 2.00 per cent sulphuric an- 
hydride has been added, the strength both in neat and mortar 
decreases with increase of the substance. Four of the cements 
show a maximum strength with approximately 1.50 per cent 
sulphuric anhydride, while two show a maximum with about 
2 per cent. The 90-day briquettes develop maximum strength 
with a higher percentage of sulphuric anhydride than do the 
briquettes for the shorter periods. Additional neat and mortar 
briquettes have been made for periods ranging from six months 
to five years. No final conclusions can be drawn until the 
long-time tests are available. In making the calculations, the 
strengths of briquettes made with the cements as received were 
taken as 100 per cent/ 4 and the comparison of results were made 
on the basis of the percentage loss in tensile strength. The per- 
centage loss in strength caused by placing plaster equivalent 
to about 10 per cent of sulphuric anhydride in each cement was 
then studied. The following points become apparent: 

1. With one exception, the percentage loss is greater with 
mortar briquettes than with the corresponding neat briquettes. 

2. As the age of the neat briquettes increases from seven 
days, the percentage loss decreases. No definite conclusions can 
be drawn from the results of the mortar briquettes at present. 

3. No relationship between the chemical composition of the 
cements and the effect of sulphate is apparent. 

INFLUENCE ON STORAGE FACTORS 

It often happens that a cement is satisfactory at the time of 
manufacture, but becomes quick-setting or develops some other 
fault during storage. These alterations may take place grad- 
ually over a period of several weeks, or in as many days. Some 
writers are of the opinion that there is some relation between 
these changes and the sulphuric anhydride content. Accelera- 
tion tests have been devised to determine in advance some of the 
changes that may take place in cement during storage. One 
such test is to spread the cement in thin layers exposed to the 
moisture and carbon dioxide of the air and test from time to 
time for soundness and setting time. 15 A test recommended 
by Hentschel 1C is as follows : Two hundred grams of cement 

11 Brand F was quick-setting as received; so a sample containing 1.46 per 
cent sulphuric anhydride was taken as the standard. 
"This Journal, Sec. A (1911), 6, 207. 
"Hentschel, G., Tonind.-Zeitg. (1912), 36, 557. 



xii. a, 3 Witt and Reyes: Calcium Sulphate on Cement 



141 



are placed in a large Erlenmeyer flask; carbon dioxide, washed 
in sulphuric acid and in water, is passed into the flask for ten 
minutes. The contents of the flask are then shaken violently. 
This process is repeated three times. If the setting time of 
the cement becomes less than one hour, it is likely to deteriorate 
during storage. The latter test was applied to a number of 
samples of brand F chosen at random and to brands A to E for 
comparison. The results are shown in Table IV. 

Table IV. — Effect of carbon dioxide absorption on the setting times of 
cements. 



Brand. 


Sul- 
phu- 
ric 

anhy- 
dride. 


Original cement. 


After absorption of CO2 
for ten minutes. 


After absorption of CO2 
for twenty minutes. 


Loss 
on 
igni- 
tion. 


Wa- 
ter 

used. 


Time of 

setting. 


Loss 
on 
igni- 
tion. 


Water 


Time of 
setting. 


Loss 
on 
igni- 
tion. 


Water 


Time of 

setting. 


Initial. 


Final. 




Initial. 


Final. 




Initial. 


Final. 




P.ct. 


P.ct. 


P.ct. 


H. m. 


H. m. 


P. ct. 


P. ct. 


H. m. 


H. m. 


P. ct. 


P. ct. 


H. m. 


H. m. 


A .... 


0.67 2.67 


24 


5 15 


8 40 


2.85 


24 


5 20 


7 40 


3.28 


24 


3 45 


6 50 


B .... 


0.86 3.07 


23 


5 10 


8 40 


3.10 


23 


5 30 


7 50 


3.75 


25 


3 45 


7 


C 


1.34 1 2.06 


25 


5 30 


8 40 


2.10 


25 


5 15 


8 5 


2.58 


25 


4 40 


8 


D..__ 


1.34 t 1.36 


23 


3 40 


6 40 


1.56 


24 


5 20 


9 50 


.1.59 


24 


5 5 


9 35 


E .... 


1.24 2.11 


23 


4 50 


8 


2.18 


23 


4 35 7 


3.11 1 23 


3 30 


6 45 


F .... 


2.00 


2.60 


22 


1 55 


3 45 


3.06 


22 


7 


3 10 


3.21 


24 


1 20 


4 10 


Fl... 


1.50 


2.13 


22 


2 35 


5 60 


2.30 


23 


1 50 


5 40 


2.37 


25 


3 5 


5 50 


F2... 


1.20 


2.89 


25 


5 5 


7 45 


3.07 


26 


4 25 


8 5 


3.16 


26 


4 35 


8 15 


F3... 


1.41 


3.80 


25 


4 10 


7 20 


4.03 


25 


4 35 10 


4.19 


26 


6 15 


9 45 



a All samples were tested for soundness and found satisfactory. 

In these examples it can be seen that there is no relation be- 
tween the amount of sulphuric anhydride present and the change 
in setting time caused by absorption of carbon dioxide. The 
sample of brand F having the highest percentage of sulphuric 
anhydride is the only one that becomes quick-setting. 

According to Hentschel 17 the change in setting time of a 
cement in storage is due to the formation of alkaline carbonates 
from absorption of carbon dioxide. The cements shown in Table 
IV were gauged with sodium carbonate solutions of three con- 
centrations. It can be seen from Table V that the set was but 
little affected. 



Loc. cit. 



142 The Philippine Journal of Science i9n 

Table V. — Effect of sodium carbonate on the setting time of cements. 



Brand. 


O 
ra 

0) 

■o 
•E 
•o 

J3 
C 

3 

£ 
"3 


Concentration of solutions. 


4 grams per liter. 


10 grams per liter. 


20 grams per liter. 


40 grams per liter. 


■ J 

»| 

e J 
SJ « 
3 u 

8*8 

o | 

si 

Is 
> 


Time of 

setting. 


|| 
> 


Time of 
setting. 


»c 

§1 

1| 

'o I 

o S 

£ 3 

II 
> 


Time of 
setting. 


.2 £ 

lb 
oi 

|s 


Time of 

setting. 


"c 


i 

£ 


c 


c 

E 


c 


c 
E 


'£ 


"3 

E 


A 

B 

C 

D__ 

E 

Fl 

F2 

F3 


0.67 
0.86 
1.34 
1.34 
1.24 
1.50 
1.20 
1.41 


cc. 
24 
23 
25 
23 
23 
22 
26 
25 


H. to. 

5 20 

6 50 
5 45 

3 50 
5 45 
2 10 
5 5 

4 55 


H. to. 

9 25 
9 50 
9 50 
5 55 
8 15 
5 45 
8 
7 55 


cc. 
24 

25 
23 

23 
22 
26 
25 


tf. m. 

5 40 
5 35 
5 30 
3 5 

3 40 
2 5 

4 10 
4 


H. TO. 

9 
9 
9 25 

6 5 

7 50 
5 35 
7 50 
7 50 


cc. 

2! 
23 
25 
23 
23 
22 
26 
26 


H. TO. 

4 20 

5 25 
5 
1 50 
4 
1 50 
4 5 
4 


fl. TO. 

8 40 
8 25 
8 40 
5 
7 30 

5 20 
7 30 

6 55 


cc. 
24 
23 
25 
23 
23 
22 
26 
26 


H. m. 

4 15 

5 
4 45 

1 

2 45 

2 15 

3 10 
3 


H. TO. 

7 45 

8 15 
8 15 

4 20 

5 30 

6 15 

7 15 
7 20 



The following statements concerning the quantity of sulphuric 
anhydride in cement are based on the cements studied in this 
laboratory for periods 1S up to ninety days : 

1. The maximum retardation of set is produced by 1.5 per 
cent to 2 per cent. 

2. In general, the briquettes show a decrease in strength when 
the cement contains more than 2 per cent, except in brands 
E and F, which decrease in strength only when the content 
amounts to about 5 per cent. 

3. The soundness (five hours in steam) is unaffected by any 
amount investigated. 

4. The percentage loss in tensile strength is greater with the 
mortar than with the neat briquettes. 

18 Measurements of expansion bars are being made over a period of 
months, and briquettes for long-time tensile-strength tests are being 
matured. Those that have been since completed indicate no well-defined 
relation between tensile strength and sulphuric anhydride content, but show 
that, in general, the maximum strength of the mortar briquettes corresponds 
to a higher percentage of this compound than in the series discussed in 
this paper. With most of the cements investigated, sulphuric anhydride, 
in quantities less than 3 per cent, caused no serious expansion in sea water. 



xii, a. 3 Witt and Reyes: Calcium Sulphate on Cement 143 

5. The percentage loss in tensile strength of the neat briquettes 
decreases as the age increases from seven days. 

6. No relation between the sulphuric anhydride content of 
a cement and the effect of exposure to carbon dioxide is apparent. 

7. Results indicate that the formation of alkali carbonates is 
not an important factor in the change of the physical properties 
of cement during storage. 



THE .RADIOACTIVITY OF PHILIPPINE WATERS* 

By J. R. Wright j and George W. Heise 

{From the Bureau of Science, Manila) 

one plate and two text figures 

In recent years there have been so many investigations of the 
radioactivity of waters that comparative data are now available 
from many parts of the world. It is the purpose of this paper 
to present the results of similar work on the radioactivity of 
typical Philippine waters. 2 

Up to the present time our work on radioactivity has been 
limited to the determination of radium emanation and actual 
radium content of typical springs and deep, drilled wells. As 
yet, no attempt has been made to study the radioactivity of the 
deposits or residues from springs, or of typical rocks, or of the 
gases evolved from springs or wells. All of the work was done 
on Luzon, most of it in Mountain and Laguna Provinces. The 
majority of the places visited are shown on the accompanying 
maps (figs. 1 and 2). 

The geography and the geology of the country in and around 
Baguio have been described by Eveland, 3 and of the remainder 
of Mountain Province by Smith.* 

Owing to the difficulties of travel, equipment for provincial 
expeditions had to be reduced to a minimum; in consequence 
comparatively few data on the chemical composition of the 
waters were secured. 

EXPERIMENTAL PART 

The testing apparatus was one of the Spindler and Hoyer 
aluminum leaf electroscopes used in previous radioactive meas- 

* Received for publication January 29, 1917. 

1 Professor of physics, University of the Philippines. 

2 The only previous attempt to determine the radioactivity of Philippine 
waters was made by R. F. Bacon [This Journal, Sec. A (1910), 4, 267-280], 
but his results were only qualitative, being obtained under conditions which 
made even approximately accurate results impossible. 

' Eveland, A. J., Notes of the geology and geography of the Baguio 
mineral district, ibid., Sec. A (1907), 2, 207-233. 

* Smith, Warren D., Notes on a geologic reconnaissance of Mountain 
Province, Luzon, P. I., ibid., Sec. A (1915), 10, 177-209. 

145 



146 



The Philippine Journal of Science 



urements 5 in the Philippines. The instrument was provided 
with the usual equipment of tripod, shaking vessel, and cir- 
culation system necessary for field work (Plate I, fig. 1). 




Fig. 1. A part of northern Luzon. 



Because of the high humidity generally encountered in the 
Philippines, a small tube of calcium chloride was fastened in- 
side the leaf chamber of the electroscope. In all determinations, 
in the field or in the laboratory, a calcium chloride tube (gen- 



5 Wright, J. R., and Smith, O. F., Physik. Zeitschr. (1914), 15, 31-39; 
This Journal, Sec. A (1914), 9, 51-77; Phys. Rev. (1915), n. s. 5, 459-482; 
(1916), n. s. 7, 49-61. 



Wright and Heise: Philippine Waters 



147 



erally of the Bender and Hobein type) was introduced into the 
circulation system, so that all air and gases were dried before 
they entered the ionization chamber. It was found impossible 




Fig. 2. South-central Luz 



to obtain reliable readings without this precaution. Our results 
indicated that drying is also necessary at lower temperatures in 
order to ensure correct results with the type of instrument 



148 The Philippine Journal of Science ion 

used. With the precautions noted, the natural leak of the in- 
strument was usually lower in laboratory work, and frequently 
lower in field work, than the figures for the natural leak fur- 
nished by the instrument makers. The electroscope was charged 
by means of a bank of storage cells in the laboratory; in the 
field, by means of an ebonite rod. 

Standardization of instrument. — The electroscope was stand- 
ardized by means of a radium bromide solution of known strength 
furnished for this purpose by the Bureau of Standards, Wash- 
ington, D. C. A definite portion of this solution was kept sealed 
in a proper vessel (essentially a "Curie tube" 6 ) for over one 
month. The emanation was then removed from the solution by 
boiling and by circulation of air and was drawn into the eva- 
cuated ionization chamber. The arrangement of the apparatus 
is shown graphically in Plate .1, fig. 2. 

A determination was conducted as follows: After the instru- 
ment had been charged, it was allowed to stand until the leak 
had become small and nearly constant. Readings both of the 
natural leak and of the leak due to emanation content were 
taken at intervals of five minutes, as measured by a stop watch, 
and were estimated to tenths of a division on the telescope scale. 

The sealed vessel B, containing the acidulated sample to be 
tested, was heated in a water bath F to about 90° C." Little 
danger of breakage was incurred by this procedure, because the 
vessels had been heated to about the same temperature before 
being sealed. The ionization chamber A was evacuated by 
means of an electrically driven Geryk oil pump D. A mano- 
meter E was attached, in order to measure the pressure in the 
chamber and to detect leaks in the joints of the tubing or in the 
ionization chamber. When a vacuum had been secured, the stop- 
cock H leading to the pump was closed. As the ends of the 
glass aspirator tubes had been drawn to a point and notched 
with a file, slight pressure with the fingers at M was sufficient 
to break off the point of the delivery tube inside the rubber 
tubing, thus allowing the vapors and gases from the vessel 
B to pass slowly through the (calcium chloride) drying tube C 
and into the evacuated ionization chamber. Under the decreased 
pressure, the liquid in the flask B boiled readily, and all of the 
emanation was rapidly drawn into the ionization chamber. In 
order to remove the last traces of emanation, the point of the 

9 Curie, P., Dosage du radium par la mesure de l'emanation degagee, he 
Radium (1910), 7, 65-70. 



xii. a, 3 Wright and Heise: Philippine Waters 149 

glass tube entering the flask was broken off at L and the flask 
was washed out with a current of air, regulated by means of a 
screw-stopcock J, until the pressure in the system had been 
equalized. The cock / was then closed, and after fifteen minutes, 
readings of the leak were taken as described above. 

The standardization was checked with a known quantity of 
standard radium-bromide solution put directly into the shaking 
vessel used in field determinations. After the closed vessel had 
been allowed to stand one month to ensure radioactive equili- 
brium, a determination was made by the usual field method. 
Two different determinations showed satisfactory agreement 
with each other and with the results previously obtained. This 
procedure is much simpler than the usual method of standardiza- 
tion, and since determinations both with the standard solution 
and with water samples of unknown activity are thus performed 
under identical conditions, the chance for error is minimized. 
Our data indicate that this method can be relied upon for ac- 
curate results. 

■The field determination of radium-emanation content. — The 
manner of taking water samples was, of course, dependent on 
the exigencies of travel and on the location and type of the 
source of the water to be examined. All reasonable precautions 7 
necessary to secure representative samples without loss of ema- 
nation were observed. Whenever possible, the collecting can 
(shaking vessel) was dipped directly into the water to be exam- 
ined; otherwise the water was allowed to flow directly into the 
can from the source (for example, a flowing well) , or was trans- 
ferred from the source in a second vessel and poured into the 
can. In all cases care was taken to prevent loss of emanation 
through shaking or aeration. 

The method of taking samples was not free from objection in 
every instance, and sometimes, for example, when a spring 
emerged from the bed of a river or in the bottom of a large 
pool, it was found impossible to obtain a representative sample. 
However, we believe that with the exception of these cases the 
error due to the method of taking the sample did not affect ap- 
preciably the accuracy of the results. 

The shaking method of Schmidt 8 was employed in field work. 
Determinations were made as soon as practicable after a sam- 

' Engler, C, Sieveking, H., and Koenig, A., Neue Beitrage zur Messung 
der Radioaktivitat von Quellen, Chem. Zeitg. (1914), 38, 425-427. 

s Schmidt, H. W., Tiber eine einfache Methode zur Messung des Emana- 
tionsgehaltes von Fliissigkeiten, Pliysik. Zeitschr. (1905), 6, 561-566. 



150 The Philippine Journal of Science m? 

pie had been taken, usually within thirty minutes, and in no 
case after more than three hours. The arrangement of the ap- 
paratus is shown in Plate I, fig. 2. 

The procedure in an ordinary field determination was as 
follows : The instrument was charged, and the natural leak was 
determined in the usual manner. The shaking can, containing 
a fresh-water sample, was then held vertically and open, until 
the water had drained to the level of the lower stopcock. The 
can was then closed, vigorously shaken for about one minute, 
and placed in the circulatory system as shown in Plate I. 
The air and the emanation in the system were circulated by 
means of a small rubber bulb for two minutes to ensure equal 
distribution of the emanation. The stopcocks on the ionization 
chamber were then closed, and fifteen minutes later the leak 
was determined, as previously described. 

Determination of radium content. — Samples of typical waters 
were brought to the laboratory, sealed in proper vessels, allowed 
to stand for a month, and tested for radium content. The 
method of procedure was identical with that described with the 
standard solution, except that 400 cubic centimeter Jena flasks, 
equipped with aspirator tubes, were used to hold the water in- 
stead of the tube used for the radium solution. About 250 
cubic centimeters of water were used in a determination. In 
some cases, when larger quantities of water were available, as 
much as 15 liters were evaporated to 250 cubic centimeters, 
acidified with ^hydrochloric acid, sealed in Jena flasks, and 
tested. 

Calculation of results. — Radium-emanation content was cal- 
culated in terms of the weight of radium that would produce 
the leak noted when in radioactive equilibrium with its own 
emanation, and is expressed as grams X 10- 12 per liter of water. 

The apparatus constant was calculated from the formula 9 

a== 1000 fr+fc+^ +ffi (1) 

in which 

w is the quantity of water, in cubic centimeters, under investiga- 
tion; 

l lt l 2 , and l. A are the quantities of air (in cubic centimeters) 
in the shaking vessel, in the respiration system, and in the 
ionization chamber, respectively; and 

a is the fraction of the emanation remaining in the water after 
shaking. 

' J Schmidt, H. W., op. cit., 565. 



xii, a, 3 Wright and Heise: Philippine Waters 151 

The radioactivity of any water expressed in grams of radium 
may be then readily calculated from the formula 

m 1 =am r7 (2) 

Z 
in which 

a is the constant derived from (1) ; 
Z is the leak caused by m grams of radium (determined with 

the standard solution) ; and 
Z ' is the leak caused by the emanation from the water in 
question. 

Since it was not practicable, especially in the field, to wait the 
three hours necessary for the leak due to emanation to reach 
maximum value, readings were taken, as previously mentioned, 
fifteen minutes after filling the ionization chamber, and the 
maximum leak was calculated by means of a radium-emanation 
decay curve. 

For use in the above equation the readings for the leak (Z 
and Z') were reduced by means of a calibration curve from 
divisions of the telescope scale to volts, and the natural leak of 
the instrument was in each case subtracted from the leak ob- 
served with emanation in the ionization chamber. 

Limits of accuracy and probable error. — Although the object 
of this work was primarily to get reliable comparative data 
rather than exceedingly accurate absolute values, it is probable 
that the error was not great. The radium solution used for 
standardization may be considered accurate within 5 per cent. 
Duplicate determinations made with the standard solution 
checked within 1.5 per cent. Duplicate determinations made 
in the field on the same water checked within the limits of ob- 
servational error. We have reason to believe that, except for 
the isolated cases noted in which proper samples could not be 
secured because of the nature of the source, the maximum error 
in field determinations was not greater than 7 per cent. There- 
fore the probable error was much smaller and hence was well 
within the limits of accuracy to be expected for this class of work. 

The radium-emanation content of Philippine waters. — No at- 
tempt was made in the field to determine anything but radium- 
emanation content. That we were actually dealing with radium 
emanation was shown by the fact that when the gases from a 
number of waters studied were allowed to remain in the ioniza- 
tion chamber for long periods of time the typical decay curve 
of the radium emanation was obtained. 

The radioactivity of the Philippine waters examined is shown 
in Table I. 



152 



The Philippine Journal of Science 






1 


Nonthermal. 

Do. 
Nonthermal, mildly sulphu- 
reted. 
Do. 
Nonthermal. 

Do. 
Large nonthermal spring. 
Series of large hot and cold 
springs. Sample from cool 
portion (31° C). 
Sample from hot portion 

(45° C). 
Temperature 23° C. 
Temperature 70° C; probably 
from same source as water 
for sanitarium. 
Temperature 22° C; probably 
seepage from a river in vic- 
inity. 
Temperature 25°C. 

Do. 
Temperature 21.5° C. 
Temperature 22° C. 
Temperature 25° C. 


A . a" 

,8 s 5 


2, 106 
(<0 

neg. 
1.284 

1.293 
neg. 

trace 

41 

neg. 

neg. 
430 

528 

1,297 
526 
317 
426 
496 I 


| 
1 

1 




| 
> 




t 

i 

£ 


e 






1 

i 

o 

c £ 
§ | 

"3 ~ 
.s 1 
8 I 

t, c 

pp > 


4 


1 

I 

I 

1 

c 


Igneous conglomerate ce- 
mented by tuff. 

Loose volcanic agglomerate- 
Igneous conglomerate 


I : a 
i II 

i' 6 

1 I o 

i ; "3 

i i 1 

ii I 

! ! c 


3 

o 
m 


| 

"< 

B 

c 

| 

< 


e 
> 
* 

1 


1 

c 

j 

< 


,2 

K 


c 


( 

E 

i 

c 

1 

< 


1 

c 
« 

< 


t 
1 

DC 
■j 

: 
P3 


e 

a 

"o c 

1 X 
Ph 


£ 

c 

V 

E 



£ 

1 
c 

m 

1 

o = 

a c 


Sinabac Spring 

San Diego Spring * 

San Vicente Spring 

Buenaventura Spring 

Small spring 


6 

I 
c 
c 

a 

_c 

£ 

0, 


( 

P 

i 

c 

1 

p 


i 


j 

c 
] 


| 
a 

a 

c* M 
3 W 

m 


pi 

a 
| 


'* 


: 
P 

i 

r! 

a 

c 

t 

(5 


1 

c 

« 
a 

a 
j 

a 
C 

l 


X 


e 

h 

a 

c 
= 


! 

II 

I 

a 
c 
j 


C 
[ 

i 

5 


i 
c 

s 

j 

'a 

i 

a 


c 
| 

i 

i 


• 3 

1 .Q j 

! 3 

: pa | 

ii 

! * a> 

; m "S 
ilfi 

! a a 
o c c 

•^ a s 

; a a 

: j ,-i 




» o> « o ^ a u £ 
S p p 3 ft 3 3 

l-S I-, I-, < h, I-, 


.0 0) O. o OOOOOI 

5: = =5 = ;pp;p 







Wright and Heise: Philippine Water, 



153 









































1 431 t 






8 


S 


















g o < 2 S c 

° 2 ft S ° 






c 
o 









o | 


. u . . o 












s»^ * -8 • 






"3 


« 




* - 


perature 31° C 
perature 33.5° 
perature 32° C 
perature 31° C 
perature 25.5° 
thermal. 


d 


|$ 


fi 




d 


Do. 

thermal; situa 
San Pablo La 
perature aboi. 
ted in bed 
ver; sample 
xed with rive 
thermal; locat 
perature 36° C 
thermal. 


d 


ft 

3 


1 


s i 


5 


3! 5 
§ § 


§ 1 1 1 i § 












g -S | S £ S g g § 




o 


o 


e o 


g 


5 a 


H H Eh Eh H Z 












Z Eh Z Eh Z 




fe 


5 


z 




■s o 


gaisss 


jj 


■a 


i 


1 


-? 


391 
448 

324 

131 
nil 
242 


9) 


'5 '£ 


c 


'5 




? 1 

i a l 
■s s ■« 

MS? 



•1 ^ -5 c to Q. o X 



g 
'ft 3 
o 3 
a 

8«- & Z % 

1 $ - c s s 



.2 a 
1-3 



a. I? 



n n S <; « (2 ? 



os ™ c S c S 

C ^ 3 S 3 'E 
W J S <J S Q 



d d 



H Ch Ph 







X 


a 













J 




c 






00 




















Ph 


Ph 










w 

a 





Ph Ph Ph CO CO 
a a a a a 

3T)333333'n3'a3 

a '• a a a a 3 a a > a ! a 



fc o 



o o 
o c c ° M 
■a J3 E> "S B 



3| * 



o-2 & 



^■8 



1 si 



SI «. S 
I o, -S •„; Z 



56- 
8 .2 » 

g -C ^ .>» 



•a c c "o c "a 



154 



The Philippine Journal of Science 



a 
1 


Nonthermal; about 1 kilo- 
meter above Headwaters 
mine. 

Nonthermal; seepage eleva- 
tion 820 meters. 

Nonthermal; seepage eleva- 
tion 1,060 meters. 

Do. 

Nonthermal; used for munic- 
ipal supply, elevation 1,250 
meters. 

Nonthermal; supplies Camp 
John Hay, elevation 1,200 
meters. 

Nonthermal. 

Nonthermal; supplies Domin- 
ican building. 

Nonthermal; located beyond 
Topside. 

Nonthermal; used for munic- 
ipal supply. 

Nonthermal. 
• Do. 


ill.' 


nil 

93 
nil 

trace 

108 

194 

122 
111 
137 
163 
trace 


c 

1 

1 

1 

O 


X 

c 
< 


< 


Andesite and diorite. vol- 
canic sinter overlying 
tuff. 

Decomposed andesite and 
diorite. 


I 

IB 

c 
'5 
H 

| l 




i 

X 

c 
< 


I 

1 

c 
| 

c 

I * 

1 i 


i 


3 


_1 

> 
C 
J 

E 
< 

b 
c 

c 

V 


Spring, Antimok Valley at 
Benguet Consolidated mine. 

Spring, Antimok Valley at 
Headwaters mine. 


X 

c 
( 

1 

(X 

\ 

I 


Camp John Hay, No. 1 spring 

Camp John Hay; spring at ice 

plant. 
Dominican Hill spring 


Government Center spring.. 

Pakdal Spring 

Sanitary camp spring 


6 

XI 

1 
a 

I 

c 
> 
8 

Oh 


t 

'5 
i 

& 
1 

c 

1 


c 


e 
x 


i 

a 

e 

1 

■ 
P 

c 

1 
c 

E 


x 


o c 


c 

X 


c 


o e 

XI X 


o 

3 

a 
a 

I 




1916. 
May 1 

May 27 
May 1 

May 27 
Apr. 27 

Apr. 29 


oj a o 
S < < 


a 

< 


6 

Z 


5 SS ™ 3" S§ S33SSS 



Wright and Heise: Philippine Waters 



155 



£ s I 

£%£ 
§ 1 § 






P ° S 
S * 1 



t S* Mao 

c e c S E £ 

o S o a § g 

Z 2 H H 



£ I 



■a § 2 

I 15 § 



5 5 



!i 



.5 H ° 

=5 .5 .£ ti 



o o o 



O i O I O I O i O O iOOO ii 

a j a j s-;a ; a a : a a a : ; 



: a 



S 3 



156 



The Philippine Journal of Science 



1 


^Temperature 50°; from dif- 
| ferent springs of a series. 

Elevation 1,660 meters. 
Sample from boiling saline 

spring; elevation 1,200 

meters. 
Nonthermal; spring about 1 

km. east of rest house; eleva- 
tion 1,350 meters. 
Nonthermal; elevation 1,650 

meters; probably seepage. 
Do. 
Probably seepage; elevation, 

2,050 meters. 
Nonthermal; water supply for 

mission; elevation 1,600 

meters. 
Elevation 1,400 meters. 
Elevation 1,200 meters. 
Elevation 900 meters. 
Tagudin artesian well waters 

are mildly sulphureted. 
Temperature about 35° C. 
Nonthermal. 


1 &* ' 


neg. 

trace 
trace 

nil 

114 

nil 

nil 
nil 

111 

trace 
263 
nil 

trace 

137 
341 


c 

o 

1 

o 
"3 

O 


|| 
o JO 

B 1-8 




"3 4J 11 

s S 1 
•3 § J 

If. S 5 

si * 

"as ; 
co 2 
<! H 


a 
< 


o 

•a 


o 


c 

E 
2 


1 - 
■o .2 

s * 

1 s 

o ™ 
«S c 
Q < 










o 
W 


1 

V 

X 

c 
c 

U 


c 


: 

c 

c 
1 

i 


C o 

■C (£ 
,J | 

c .! 
1 1 


< 
i 

i 

"1 

I £ 


c 
| 

b 

c 

1 c 


; c 
i 1 

E 

! 1 


'1 
| 

1 


|| 
^ fc 


Artesian well, plaza 

Artesian well; Calle Espanol 

Dismantled artesian well near 
cemetery. 


"E 

a 
1 
a 
| 

c 


t 

J 

T 

1 

u 

J 

] 
i 

S 


< 


I 

i 


1 

i 

'I 

c 

1 c 


i 

! 

1 

t 
i 

si 


s 

c 

! I 

B § 
i 1 J « 

1 ?r 

3 a 


2 

"3 

c 
B 

C -5 
3 a 
O b 

i « i 

3 « i 

\ § £ ! 

1 £ 


i 

< 


; 

E 

■ 


I c 
1 


] 

■i 
i 
E 
[ 

i 


1 
4 


> e 


• 





H S 


. d d rt d 

: S s S s 


5 rt < 

; & ■& 
: a a 


a 




> < 

] 7 


> 


6 


ea co •** w to t- oo cs o r* oa co ** 10 (O 
c- r-t-c- t- c- t- t> oo aooooooo oooo 



Wright and Heise: Philippine Water i 



157 



a 

t ii 

.E -o 

u >> o 

£ -° £ 

s 15 

a I 1 

>, 

3 "good 

5 a c Q Q Q 

e a ■= 

ft. tc 

I i Mi 

1 i i i. i 

■ 1 i : i 
2 . : : : 

° w : : : 

I ^ 11 i 

I °. IM 
! i ill 

s « 5 , 5 S S 
to "3 -2 g .3 -| | 

'jj •§ £ S t t t 
c 

I Q i i g 

h c : i •§ 

= c : : S 

1 '1 o o * 
§ c ^ "°. 1 

s (2 i : « 

o [J j i es 

as : : 4 
£ So So § s 



158' The Philippine Journal of Science mv 

Radium content of Philippine waters. — Samples of twenty 
typical waters, from which the emanation had been removed, 
were acidified, sealed up in proper containers, and allowed to 
stand at least one month. They were then tested for radium 
content, by the method previously outlined. In no case was there 
any indication of radioactivity. 

In addition, 15 liters of Batangas water (No. 1), 15 liters 
of Los Banos water (No. 12), and 5 liters of Sibul Springs water 
(Nos. 4 and 5) were evaporated to small bulk and similarly 
tested. The first two showed no emanation, and the third 
showed a trace. 

The radioactivity of the waters studied was, therefore, 
primarily due to emanation derived from the materials in the 
ground with which the water had been in contact and not to 
dissolved radium salts. 

DISCUSSION 

The work has not proceeded sufficiently to justify many con- 
clusions. The typical Philippine water supplies studied possess 
no abnormal features so far as their radioactivity is concerned. 
Though some of them are moderately high in radium-emanation 
content, none show an excessive amount, compared with waters 
from other countries reported in the literature. 10 

Since hot water is a poorer solvent of gases than cold water, 
it is to be expected that the radioactivity of hot springs should, 
in general, be low. With the exception of the Los Banos water 
(No. 12), most of the thermal waters studied in the course of 
this work showed little or no activity. 11 

In general, the average activity of igneous rocks is greater 
than that of the sedimentary, 12 and it is to be expected that water 
from the former material should show higher emanation content. 
Thus Sahlbom 13 found that the water from sedimentary deposits 
was much lower in activity than that from primary rocks; 
further, that wells bored in the acid rocks showed the highest 
activity. In the Philippines the relatively small number of de- 

10 Cf. Schlundt, H., Bull. U. S. Geol. Surv. (1909), No. 395, 31; Jown. 
Phys. Chem. (1914), 18, 662. Isitani, D., Proc. Tokyo Math. Physic. Soc. 
(1912), and following years. 

11 The average temperature of ground waters in the lowlands of Luzon 
is about 28° C. 

12 Cf. Clarke, F. W., Data of geochemistry, Bull. U. S. Geol. Surv. (1916), 
No. 616, 122. 

"Sahlbom, N., Arkiv Kemi, Min. Geol. (1915), 6, No. 3, 1-52; through 
Chem. Abst. (1916), 10, 1134. 



xu. a. s Wright and Heise: Philippine Waters 159 

terminations made and the frequent difficulty of determining the 
actual water-bearing stratum, since this is frequently not the 
same as the geological formation exposed at the place where 
the water emerges, make generalizations at this time inadvisable. 

In some cases, at least, the radioactive material from which the 
water derived its activity must have been confined to a rather 
limited area. Thus it was pointed out ' ' that Olla Springs 
(No. 13) were in reality only seepage water derived from a 
river about a hundred meters distant. 1 '' The "spring" water 
must, therefore, have acquired its activity in the course of a 
short journey underground. 

So far as the available analytical data at hand are concerned, 
there is no apparent general relation between the chemical quality 
of the water and its radioactivity. According to Schlundt in 
there is, in general, no detectible difference in activity between 
acid and alkaline waters. Practically all of the waters tested 
at the source in the course of this work were acid to phenolph- 
thalein and alkaline to methyl orange. We have been equally 
unable to make generalizations concerning other factors. This 
is not surprising, since the emanation content appears to be 
due not to dissolved radium but to contact with radioactive 
materials, sometimes within a very restricted area. For the 
sake of completeness, the available analyses of the waters under 
investigation, as compiled from the data in the Bureau of 
Science, are included in Table II. 

There was no sharply defined rainy season 1T during 1916 
in the places visited. With the exception of the determinations 
made in April and December, the tests for radioactivity were 
conducted during months of considerable rain. 

Though no systematic study of the relation between the radio- 
activity and the variation of flowing wells and springs has been 
made, 18 it may be of interest to point out that Sibul Springs 
(Nos. 4 and 5) was tested on two different occasions — once in 
the middle of the dry season, that is, after two or three prac- 

" By V. E. Lednicky, chief, division of mines, Bureau of Science. 

15 A bacteriological test made in the course of a field survey of water 
supplies confirmed this view. 

"Schlundt, H., Bull. U. S. Geol. Surv. (1909), No. 395, 30. 

17 For the distribution of rainfall in the Philippines according to locality 
and season, see Cox, A. J., This Journal, Sec. A (1911), 6, 287-296. 

1S Many spring's and deep wells whose flow varies greatly with the tide 
show no appreciable variation in chemical quality. See Heise, G. W., Note 
on the tidal variation of .springs and deep wells in the Philippine Islands, 
This Journal, Sec. A (1916), 11, 125-127. 



160 



The Philippine Journal of Science 



(>OS) S3}Btt<IltlS 



3 2 Z 8 



•(sOOIDs^'Joq-^ia 



lO C- N t- C- CO tO C- lO »-4 



•(CQO) EB^EUOqaBO 



Z Z Z22ZZZZZZ Z ZZEh 



•(10) snuomo 



» ft 



■8 3 

2 g 



to a 

"> V. 
3> II 



.8 II 
g w 



•(3N) uimsaoSEM 



•(BQ) UiniOIEQ 



o © © ^ o o o 



•spnos [KIOJ, 



•(300) *»!PPV 
•(EOO^O) ^.'UH^siIV 




•a^nuiui jad ^ioecIbo 



« Jl.3 a -3 3 

. S co w ffloS 



8.S 



'3 Z 



SSS 3 8 



III 



a to g 



0< Oh 



5 « « a 

M 3 3 « 

05 M 03 »J 



889 
j j j 






2 5 



CO CO CO 



■Hi 

c 3 ^ 
CO M <! 



xii. a, 3 Wright and Heise: Philippine Waters 



161 



s£ S H S 2 3 H S °> H H H H § § z ss ^ a g 


1 III 1 § || * s | m | 


S ^2 s § S3 


H HZZ z 2^ 5 Z Z Z Z Z § Z °^ ^ Z Z 


^■i a a ■* cjuso-* t- »o o " 


S | | S 5 | | 


00 










a n 










<=> H H 




c3 H 


d 




53 




J OT 










mm °° 




5? § 


d Z o 


S S 2 (M tj. 

d d £ • rH d Z H H Z Z 




22 z *! '• 

Z Z ^ o o 


s 










as 










o d o 




§ s 


1 


►J w J 


SI 














1 




1 1 


z z ~ N 


z 


+ + + + + 


+ Z Z Z Z 


~ 10 © © o o o 


5 S "- S 




S " S § ? 








I 




i 




£ 


















o = 

■s j- x; 
35 55 


z 




§ 






tf tf * tf 

3 d | rt 




1 1 




c 

c 

1 1 

ft. c 
. b 

« .2 « 

§ § I 


I 

r 

a 
c 
E 
! 


2 1 r 

Z ii 

s « * 

llsi 

3 -9 3 >- 




£ 

D 

K 

C 

.c 
'c 


a 

.5 c 
a '= 

w -8 .E 

M '5 £ 

11 = 


i 

o 

a l 

M r 

c § c 

5 « | 

C C c 

IN 


ft 

s •= 
<3 1 

ce 

>, f 

c >S c 
5 c 1 

C J3 E 
3 O e 

1^1 


1 
1 

c 

z 

- c 

J 


3 

_0 

3 .E 

tO T 

C E 

c" c 

!|| 

o -a E 


c c c 
ft ft ft 

3 3 3 

Q Q Q 

• i - ■ . i 

c o c c o c 

5 3 1 <& 1 1 1 

a % a A! a! >h ft. 

»-3 * « tc ?! ^ 

C Q 3 £ C -c « 
CM ft, W ft. W ft 




p 


6 

1 


1 

1 

1 
t/ 


C 
i 


.! 
i 


l 6 


i 

i •! 


1 
c 


1 

c 

1 g! 

c .S c 


ft 

I 


b 

I 

ft 


j 
i 

1 


ft " 
o -j 

5 -j 


c! 


p 


c 


a 





162 The Philippine Journal of Science 1917 

tically rainless months, and once during a period of frequent 
rains. The results of the two determinations, which are prac- 
tically identical, indicate that deep-seated springs, such as Sibul 
Springs, may show surprisingly small seasonal variations in 
radioactivity. It should be pointed out that Sibul Springs 
shows a comparatively slight variation in flow throughout the 
year, so that the inference from the two isolated determinations 
just discussed is not considered to be at variance with the find- 
ings of other investigators. Thus Ramsey 10 found a greater 
emanation content in certain springs during periods of wet 
weather and great flow, and Steichen 20 observed an increase in 
activity in certain Bombay hot springs during the dry season, 
while the flow of water was considerably reduced. As pointed 
out by the latter writer, local conditions may well account for 
the differences noted. 

For the sake of completeness we have noted in our data (Table 
I) all sources of water supply which are popularly considered 
to have special medicinal virtues. Either popular opinion is a 
poor guide to the medicinal value of a water, or else the medicinal 
properties of water are not to any great extent due to radium- 
emanation content. The waters with perhaps the greatest reputa- 
tion, namely, Los Banos (12) and Sibul Springs (4 and 5), have 
relatively high radium-emanation contents, yet many others 
regarded as highly, such as Marilao (3), Pansol (9 and 10), 
Santo Tomas (36) and Klondike (72), contain little or no ema- 
nation. Moreover many waters high in activity, such as Batan- 
gas (1), Kiangan (66-71), Pagsanjan (19 and 21), and in 
general, the spring waters near Mount Banajao (13-17), are 
regarded with entire indifference. 

Although emanation taken into the stomach is probably dif- 
ferent in effect from that taken into the lungs, the following 
comparison may be of interest: Assuming that in ordinary 
respiration the average human being breathes 7 liters of air per 
minute, or 10.1 cubic meter per day, the emanation content thus 
brought in contact with the human system is 770 XlO" 12 curies, 
if the normal emanation content of the air in the Philippines 21 
be taken as a basis for calculation. Therefore a person would 
have to drink about three-fourths liter of Sibul Springs water. 

19 Ramsey, R. R., The variation of the emanation content of certain 
springs, Phil. Mag. (1915), 30, 815-818. 

20 Steichen, A., The variation of the radioactivity of the hot springs at 
Tuwa, ibid. (1916), 31, 401-403. 

a Wright, J. R., and Smith, O. F., loc. cit. 



xii. ,\, s Wright and Heise: Philippine Waters 163 

or one and one-half liters of Los Banos water, in order to take 
as much emanation into his system as he ordinarily secures by 
ordinary daily respiration alone. 

SUMMARY 

The radioactivity of about ninety different Philippine waters, 
chiefly from springs and flowing wells, has been studied. The 
highest radium-emanation content encountered in a deep-well 
water was equivalent to 2100 XlO 12 grams of radium; the 
highest in a spring water was equivalent to 1300 XlO -12 grams 
of radium. 

A test for the actual radium content of about twenty typical 
sources showed that the radioactivity encountered was due to 
emanation absorbed from materials with which the ground 
water had been in contact and was not due to dissolved radium 
salts. One sample of water showed a scarcely detectible trace 
of activity due to radium salts in solution ; all the others tested 
gave negative results. 



ILLUSTRATIONS 

Plate I 

Fig. 1. Field outfit for testing the radioactivity of waters. 

2. Laboratory apparatus for testing the radioactivity of waters. 

TEXT FIGURES 

Fig. 1. Map of a part of northern Luzon. 
2. Map of south-central Luzon. 

165 



PUBLICATIONS FOE SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPFINE ISLANDS — Continued 



BOTANY 

A FLORA OF MANILA 
By Elmer D. MntRtTJ, 

Order Ho. 419. Paper, 490 pages, $2.50. 
postpaid. 

Practically a complete flora of the cul- 
tivated areat in the Philippines. Descrip- 
tions, with keys, of over 1,000 speoles, 590 
genera, and 136 families, with native names, 
glossary of technioal terms, etc. 



PHILIPPINE DLPTEHOCARP FORESTS 
By William H. Brown and Donald H. 



Order No. 432. Paper, 150 pages, 1 map, 
13 plates, and 12 diagrams, $1.00, 
postpaid. 

In Philippine Dipterooarp Forests tha 
authors present a very comprehensive discus- 
sion of the growth and development of dip- 
terocarp treea and of the other elements of 
lowland Philippine forests. 



INDO-MALAYAN WOODS 
By Fbed vV. Foxwobthy 

Order No. 411. Paper, 182 pages, 9 

Plates, $0.50, postpaid. 

In Indo-Malayan Woods, Doctor Fox- 
worthy has brought together a large amount 
of accurate information concerning treea 
yielding woods of economic value. 



ZOOLOGY— Continued 

A MANUAL OF PHILIPPINE BIRDS 

By Richabd C. McGregor 

Order No. 103. Paper. 2 parts, 769 

pages, $4, postpaid. 

A Manual of Philippine Birds oonta'ns 
in compact form descriptions of all the 
known speoles of Philippine birds. The usual 
keys and diagnoses of orders, families, and 
genera help tha novice In identification. 



By David Stars Jordan and Robert Eakl 
Richardson 



Paper, 78 pages, $0.75, 



This list will be found a convenient guide 
to the synonymy of Philippine ichthyology. 
The nomenclature i* thoroughly revised, and 
the distribution of each species within tha 
Philippine Islands is given. 



By W. Schultzb 



A LIST OF MAMMALS OF THE 
PHILIPPINE ISLANDS, EXCLU- 
SIVE OF THE CETACSA 

By No Holusiejs 

Order No. 418. Paper, 64 pagei, $030, 



The distribution Of each speoies is glv 
and the original descriptions are cited. 



Order No. 436. Paper, 198 pages, $1.00, 
postpaid. 

This catalogue Inoludes the name* of all 
species of Coleoptera that have been recorded 
from a definite locality In the PhlHppfna 
Islands. References to original descriptions 
and other Important notes are given. The 
economio appendix Includes comment on 
those species of beetles which are known to 
be Injurious or beneficial to man. 



PRICES ARE IN UNITED STATES CURRENCY 

Orders for these publications may be sent to the BUSINESS MANAGER, 
PHILIPPINE JOURNAL OF SCIENCE, BUREAU OF SCIENCE, MANILA, P. I., 
or to any of the agents listed below. Please give order number. 



The Macmillan Company, 84—86 Fifth Avenue t New York, V. S. A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinns Nijhoff, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdinandstrasse 2, Berlin N. W., Germany. 

Kelly & Walsh, Ltd., 32 Raffles Place, Singapore, Straits Settlements. 

A. 31. & J. Ferguson, 19 Balllie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 






CONTENTS 

WELLS, A. H. Destructive distillation of Philippine woods Ill 

BRILL, HARVEY C, and ALINCASTRE, CECILIO. The pos- 
sible maximum vitamine content of some Philippine vege- 
tables 127 

WITT, J. C, and REYES, F. D. The effect of calcium sulphate 
on cement 

WRIGHT, J. R., and HEISE, GEORGE W. The radioactivity of 
Philippine waters 



u. s. 

Currency. 

$2.00 

3.00 

2.00 

2.00 

5.00 

7.00 

.50 



The "Philippine Journal of Science" is issued as follows: 
Section A. Chemical and Geological Sciences and the Industries- 
Section B. Tropical Medicine 

Section C. Botany 

Section D. General Biology, Ethnology, and Anthropology (Sec- 
tion D began with Volume V) 

Entire Journal, Volume II, III, IV, or V 

Entire Journal, beginning with Volume VI 

Single numbers (except of Volume I) 

Each section is separately paged and indexed. 
Authors receive lOO copies of their papers free. 
Volume I, 1900 (not divided into sections) and supplement, sold 

only with a complete file of section A, B, or C 10.00 

Supplement to Volume I (botany) S.50 

Volume I (without supplement) , sold only with a complete file of 

section A, B, or C 6.50 

Single numbers of Volume I "5 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, P. I., or to any of the agents 
listed below: 

AGENTS 

The Macmlllan Company, 64- 8 6 Fifth Avenue, New York City, TJ. S. A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinus Nijbon*. Larige Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdinandstrasse 2, Berlin, X. "W., Germany. 

Kelly & Walsh, Limited, 32 Raines Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 19 Baillie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



Entered al the pott offioe at Manila. P. I., as »eoond-cla» matter. 



i £ : r ■ r ■ 



Vol. XII, Sec. A, No. 4 



July, 1917 



THE PHILIPPINE 

JOURNAL OF SCIENCE 



ALVIN J, COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D.; 4. R. WRIGHT, Ph. D.; G. W. HEISE, M. S. 

J. C. WITT, Ph. D.; T. DAR JUAN, A. B.; A. H. WELLS, A. B. 

R. C. MCGREGOR, A. B.; H. E. KUPFER, A. B. 




MANILA 
BUREAU OP PRINTING 

1917 







GE, SALE BY THE BUREAU OF SCIENCE, 
JKANLLA, PHILIPPINE ISLANDS 



ET1IXOLOGY 

A VOCABULARY OB THE IGOROT LAN- 
GUAGE AS SPOKEN BY THE 
BONTOC IGOROTS 



The vrjpabulary i 
i EngliSh-lgorot 



ETHNOLOGY— Continued 



By Najeeb M. Salseby. 

Order No. 405. Paper, 107 page*, 

plates, 5 diagrams, $0.25; half r 
rocco, $0.75; postpaid. 

volume deals with the earli 
-scords of the Moros In Mfndar 
The names of the rulers of Magindanao 
recorded in five folding diagrama. 



and 

TEE BATAKS OF PALAWAN 

By Edwakd Y. Mjllkr 

Order No. 403. Paper, $0.25; half mo- 

rocco, $0.75; postpaid. 
The Nabaloi Dialect (65 pages, 29 
plates) and the Bataks of Palawan (7 
pages, 6 plates) are bound under one cover. 



THE BATAN DIALECT AS A MEMBER 

OE THE PHILIPPINE GROUP 

OF LANGUAGES 



NEGRITOS OE ZAMBALES 

By William Allan - Reed 

Order No. 402. Paper, S3 pages, 62 
plates, $0.25; half morocco, 50.75; 
postpaid. 
Plate3 from photographs, many of which 
ware taken for this publication, show orna- 
ments, houses, men making fire with bamboo, 
bows and arrows, dances, and various typ*a 
of the people themselves. 



By Otto Soheerer 
and 



THE SUBANUNS OE SINDANGAN BAT 

By Emerson B. Ohm 

Order No. 410. ' Paper, 121 panes, 1 
map, 29 plates, $1.25, post,". 

igan Bay is situated on the north- 
cm Ci>.ist of Zamboanga Peninsula. The Su- 
banuns of this region were studied by Mr. 
Christie during two periods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanune at 
work and at play: 
altars, grid 
themselves. 



THE SUGAR IICDUSTRY IN THE 

I.-I.A;rD cr NZOXOS 



lo. 412. Paper, 145 

rnap, $1.25, posl 



the 



from the viewooin. 
hi. Walker's Sugar 
■f Negros is one of the most 
d by the' Bureau 
lis volume is a real oontribu- 
>ject; it is not a mere com- 
e author was in the fiold and 
conditions of which he 



E HISTORY OE STJLTJ 

y Najeeb " 

406. Paper, 275 pages, -4 
i, 2 diagrams, $0.75, pos. 

;cript for 



LIPPINE SILK 

CULTURE t 



i. 413. Paper, 53 



THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 

Vol. XII JULY, 1917 No. 4 

THE PHYSIOLOGICAL ACTIVE CONSTITUENTS OF CERTAIN 
PHILIPPINE MEDICINAL PLANTS: II 1 

By Harvey C Brill and Albert H. Wells 

(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

FOUR PLATES 



Lophopetalum toxicum. Toaldalia asiatica. 

Erythrophloeum densiflorum. Lunasia amara. 

Quisqualis indica. Rourea erecta. 

Tylophora brevipes. Hymenodictyon excelsum. 

INTRODUCTION 

An investigation of the physiologically active principles of the 
medicinal plants occurring in the Philippine Islands was begun 
by Bacon 2 in 1906. His investigation included Alstonia schol- 
aris R. Br., Datura fastuosa Linn., Caesalpinia sappan Linn., C. 
bonducella Flem., Entada scandens Benth., Tinospora crispa 
Miers., Argemone mexicana Linn., Erythoxylon burmanicum 
Griff., Aleurites moluccana Willd., A. trisperma Blanco, and 
Jatropha curcas Linn. 

It is recognized that many of the plants growing in the Archi- 
pelago that are reputed to be physiologically active have thera- 
peutic value, but there is reason to believe that certain others 
possess no medicinal properties. Many of the plants found in the 

1 Receive for publication February, 1917. Botanical material and scien- 
tific and native names were furnished by Mr. E. D. Merrill and Dr. Leon 
Ma. Guerrero, of the section of botany, Bureau of Science; the historical 
data regarding local uses of the species investigated was supplied by Doctor 
Guerrero. 

2 Bacon, R. F., This Journal (1906), 1, 1007. 

150276 167 



168 The Philippine Journal of Science 1917 

Philippines are not indigenous, but are found throughout the 
Malay Archipelago and India, while others are closely related 
to those occurring elsewhere. In India various species have been 
used both by European and native practitioners for medicinal 
purposes and are rated in their pharmacopoeias as valuable drugs. 
This investigation of Philippine plants will show whether the 
same medicinal value is possessed by the species found here, 
will classify the active principles present, and will, where pos- 
sible, furnish an accurate identification of the therapeutic body. 
No attempts will be made to furnish complete physiological tests 
of the active substances. 

Individuals of related species or even plants of identical species 
may have different physiological constituents. Chernoff and 
others 3 state : 

It is possible that saponins from the same species of plants may differ, 
depending on the place where grown and the time of year when picked. 
We have some evidence of this point from work done recently on other 
species of plants. 

The general methods used for the analysis of the plant ma- 
terials are those of Dragendorff, 4 Robert, 5 van Rijn,' ; Allen, 7 and 
other investigators who have worked on various medicinal plants. 
References to the latter are given in place. 

In our investigations work has been based on material taken 
from fresh specimens collected and identified by the botanists 
of the Bureau of Science. In order to eliminate any possible 
loss due to decomposition, volatilization, and other factors during 
the drying period, work was done on both fresh and dried ma- 
terial. Quick drying of samples was obtained in a Freas oven 
at a temperature of 101° C. No appreciable changes were found 
in the analysis of the two differently prepared samples. Before 
extraction, the dried portion of the plant, supposed to contain 
the active principle, was ground to a fine powder. After mac- 
eration or percolation in the cold from eight to ten days the 
sample was usually extracted hot with the same solvent in order 
-to assure complete extraction. Small samples of the plants 

3 Chernoff, L. H., Vichoever, A., and Johns, C. 0., Journ. Biol. Chem. 
(1917), 28, 438. 

4 Dragendorff, G., Plant Analysis. Bailliere, Tindall and Cox, London 
(1884). 

5 Robert, R., Beitrage zur Renntnis der Saponinsubstanzen f iir Natur- 
forscher, Arzte, Medizinalbeamte. Ferdinand Enke, Stuttgart (1904). 

"van Rijn, J. J. L., Die Glykoside. Gebriider Borntraeger, Berlin (1900). 
7 Allen's Commercial Organic Analysis. Blakiston's Son & Co., Phila- 
delphia (1913). 



xii, a. 4 Brill and Wells: Medicinal Plants, II 169 

were always submitted to preliminary extractions for the pur- 
pose of determining the effect of the solvent. 
- Petroleum ether was used to remove any plant paraffin, wax, 
and oil; the extract was tested in the usual way for alkaloidal 
substances and, when thought necessary, for aromatic resins and 
essential and fixed oils. The residual dry plant matter was then 
extracted with ether, and this extract was tested in the usual 
manner for the presence of alkaloids and other principles ; it was 
then allowed to evaporate, and the resinous mass was examined. 
After the extractions with petroleum ether and ether, the plant 
residue was divided into several portions to be used separately 
for the determinations of alkaloids, glucosides, saponins, certain 
oils, albumens, and bitter principles. 

LOPHOPETALUM TOXICUM LOHER (CELASTRACE^}) 

Abuab, batingui, dalinding (T. in Rizal), dayandag (T. in 
Mindoro), lanitan (V. in Samar), sudcad (V. in Masbate), 
puti-i-lalaque, puti-i-babae (M. in Lanao), and buyun (M. in 
Zamboanga) . 8 

Lophopetalum toxicum, of the Celastracea?, is a tree well 
known among the mountain people of some parts of the Phil- 
ippine Islands, due to the use they make of its bark to poison their 
arrows, spears, and other weapons. 

According to the Remontados and Negritos of Rizal Province 
the poison is easily prepared. The method is as follows: The 
bark is removed, soaked, and bruised. The expressed juice is 
then evaporated to the consistency of an emulsion. In its prep- 
aration care should be taken not to let it come in contact with 
any sour substance in order not to diminish or neutralize its 
deadly action. 

As an antidote against its poisonous effects the natives use a 
plant called tamauyan (Strombosia philippinensis Rolfe), or the 
fruits of catmon (Dillenia philippinensis Rolfe), or any other 
sour substance, like vinegar. 

The fact that the properties of the compound are changed by 
warming with acids is proof of the glucosidal nature of the com- 
pound. The natives always concentrate the juice, as they believe 
it is harmless when in the dilute condition. The change taking 
place in the dilute juice is probably a hydrolysis to a harmless 

s Throughout this paper T. stands for Tagalog; V. for Visayan; II. for 
Ilocanoj'Sp. for Spanish; F. for Filipino; Pang, for Pangasinan; Pam. 
for Pampanga; N. for Negrito; M. for Moro; B. for Bicol; and Ifg. for 
Ifugao. 



170 The Philippine Journal of Science mm 

sapogenin by the action of various ferments. Heating destroys 
these, and as yeasts will not grow in the concentrated juice, 
there are no further changes after concentration. 

Loheiy who was the first to find and describe Lophopetaium, 
at the end of his phytographic diagnosis writes the following: 

Cette espece, voisine du Lophopetaleum ftmbriatum Wall., habite les 
forets de l'interieur de l'ile de Luzon. Les Negritos, race indigene des 
isles Philippines, recolectent l'ecorce tres veneneuse de cet arbre pour en 
preparer un extrait servant a empoissonner leurs fleches. 

CHEMICAL EXAMINATION OF LOPHOPETALUM TOXICUM 

Trie fresh bast fiber was quickly dried, pulverized, and mac- 
erated with petroleum ether for eight days. Slow evaporation 
of the petroleum ether leaves a mass of a yellow crystalline sub- 
stance that is easily washed free from color by cold absolute al- 
cohol. The white residue crystallizes in drusy clusters of 
acicular crystals. This crystalline mass is odorless and tasteless 
and has no physiological effect when administered in oil to a 
guinea pig. The yield is 0.8 per cent, the melting point is 183° C. 
It is fairly soluble in petroleum ether, ether, warm alcohol, and 
chloroform, but is insoluble in water. This white substance has 
the appearance of wax, responds to no tests for nitrogen, shows 
no acid or basic properties, and is slightly acted upon by alcoholic 
potash solution. The yellow coloring matter is easily taken up 
by animal charcoal. Ether extracts from the dry residue a small 
amount of yellowish resin of an acid nature. Washing out the 
petroleum ether and ether extracts with acidulated and alkaline 
water gave no traces of alkaloids present. 

A water extraction of a portion of the plant residue after the 
above macerations froths greatly upon agitation and after 
hydrolysis reduces Fehling's solution. 

The remaining plant material was submitted to the method of 
Robert 10 for isolating saponins. The purified saponin substances 
were evaporated to dryness under reduced pressure, redissolved 
in 85 per cent alcohol, and allowed to stand several weeks in a 
vacuum desiccator. Irregular crystals separated out. A small 
amount of these was left after washing freely with 98 per cent 
alcohol. A solution of the crystals gives a precipitate with 
barium hydroxide and is easily hydrolyzed by strong acid. 
Mecke's reagent gives a purplish red coloration. 

A guinea pig, weighing 523 grams, injected intraperitoneally 
with 0.0021 gram of these crystals, died in one hour and forty- 
five minutes. Symptoms of uneasiness began after three min- 

D Loher, A., Icones Bogev (1897), 1, 55. "Op. cit. 



xii, a, 4 Brill and Wells: Medicinal Plants, II 171 

utes, followed by apparent blindness, increasing weakness of 
limbs, abdominal convulsions, exhaustion, and death. Similar 
injection into a guinea pig of 0.235 gram of the crude saponin 
substance, recovered from an extraction by methyl alcohol, gave 
the same symptoms, and death resulted after twenty minutes. 

The sapogenin obtained by hydrolysis with emulsin formed 
groups of thick acicular crystals quite insoluble in water. The 
usual methods for detecting alkaloids failed to show the presence 
of alkaloids in the plant material. An extraction of the fresh 
bast fiber in sodium chloride solution gave no nitrogen substances 
precipitated by magnesium or sodium sulphate, thus indicating 
the absence of any of the common albumens of a toxic nature. 
As sapotoxins have slight value medicinally, no further exami- 
nation was made of the compound. 

This plant contains a physiologically active substance, a saponin 
which is poisonous in small quantities. 

ERYTHROPHLOEUM DENSIFLORUM (ELM.) MERRILL (LEGUMINOS^E) 

Malatabigui, calamantao (T. in Tayabas), malabunao (B. in 
Camarines and Zamboanga), ngiricngic, albihal (N. in Cagayan), 
and salsal (Ibanag in Cagayan). 

Erythrophloeum densiflorum Merr., of the Leguminosse, is not 
considered poisonous among the Filipinos who make use of the 
wood for building purposes. 

Lewin J1 says: 

Every species of Erythrophloeum is more or less rich in erythropleine, 
an alkaloid which is reputed to be a heart poison. 

CHEMICAL EXAMINATION OF ERYTHROPHLOEUM DENSIFLORUM 

Erythrophloeum densiflorum was primarily selected for anal- 
ysis to determine whether or not it contained the alkaloid, ery- 
throphleine, which is found in Erythrophloeum guineense. The 
routine method for alkaloids showed none present. The two 
methods 12 outlined by Gallois and Hardy were also closely fol- 
lowed. Neither method gave any indications of alkaloids, either 
amorphous or crystalline. In the Stas method the tartaric acid 
was found uncombined with any of the plant constituents. When 
treated with water, the concentrated alcoholic extract leaves a 
large quantity of red powder, probably decomposition products, 
such as phlobaphenes and dextrose resulting from the action of 
the dilute acid on the tannins. Neither the methyl alcohol nor 

11 Lewin, L., Traite de Toxicologic Octave Doin, Paris (1903), 654. 
"Gallois, N., and Hardy, E., Journ. Pharm. et Chimie (1876), 24, 26. 



172 The Philippine Journal of Science mx 

van Rijn's methods for isolating and testing saponins showed 
any positive results. No crystalline products of hydrolysis were 
obtained. Several injections of the aqueous solutions were made 
into guinea pigs, but without any results to indicate the presence 
of physiologically active constituents; also the water-insoluble 
portions were administered in oil without results. Tannins are 
present, but as they show an indifferent nature in regard to 
this work, their classification was not determined. From the 
water extracts they are entirely precipitated by lead acetate, 
while gold chloride gives a positive color in the Seyda test 13 for 
tannins. 

The red substance insoluble in water and ether is fairly soluble 
in alcohol and in dilute ammonia, thus closely resembling the 
anhydrides of tannic acids or the phlobaphenes. The aqueous 
solution, from which this substance separates, strongly reduces 
Fehling's solution. 

The bitter taste found in Erythrophloeum densiflorum may be 
attributed to the presence of the natural tannins. The important 
chemical difference between Erythrophloeum guineense and the 
Philippine species, Erythrophloeum densiflorum, is due to the 
presence of the alkaloidal substance erythophloeine in the bark 
of the former plant and its absence in the latter. 

This result clearly shews that plants of the same morphologic 
characteristics may differ greatly, due to the nature of the sub- 
stances that they elaborate in their economy. 

QUISQUALIS INDICA LINN. (COMBRETACE^) 

Pinones (Sp. and F.), niogniogan (T. in Rizal, Bataan, Min- 
doro), tangolo (T- in Tayabas and Camarines), talolon (T. and 
V.), talulong (T. in Marinduque), tangolong, tangulong (T. in 
Rizal and Manila), tagarao (T. in Rizal), babebabe (Pam. in 
Pampanga), bo-nor (V. in Mindoro), balitadhan (V.), tal-lolang 
(II. in Camiguin Island), tartaro (II. in Ilocos), tahdong (Ibanag 
in Cagayan), taungon (Butuan in Agusan), and twtaraoc (Zam. 
in Zambales). 

The fruit of Quisqualis ind.ica, called niogniogan or pinones in 
Manila, is a popular remedy. It is used in every country where 
it grows, and its anthelminthic properties are widely known- 

Mercado, 14 writing in the last third of the seventeenth century, 
speaks of the anthelminthic properties of pinones seeds. Lou- 

" Seyda, A., Chem. Zeitg. (1898), 22, 1085. 

" Mei-cado, I., Libro de Medicinas de Esta Tierra, in Blanco, Flora de 
Filipinas. 3d ed. (1880), 4, 53. 



xn. a. i Brill and Wells: Medicinal Plants, II 173 

reiro " considered that the seeds of Quisqualis had not only ver- 
mifugal properties, but also that it was a corroborant astringent 
and a nephritic. 

Hanbury 10 speaks of this drug while referring to a work of 
Dr. E. J. Waring concerning the indigenous anthelmintics of 
India and remarks about its effectiveness against worms in 
children. 

Soubeiran " says that the fruits of Quisqualis are an excellent 
vermifuge if four or five seeds are taken before breakfast on the 
morning of the first days of the month. Unsound fruits or those 
that have begun to open should not be taken, as they cause 
hiccough. 

This symptom, which as a rule occurs only when too many 
of the seeds have been taken, does not occur in every individual. 
Some individuals can swallow great quantities of them without 
causing hiccough; for these it acts only as a laxative. 

It is claimed that the roots also possess the same vermifugal 
qualities as the kernel. 

CHEMICAL EXAMINATION OF QUISQUALIS INDICA 

The leaves, roots, and fruit of Quisqualis indica were analyzed. 

We found no compounds in the leaves or roots which would 
seem to have any physiological effect. No alkaloid was found. 
None of the extracts showed any results with dogs or guinea 
pigs when administered intraperitoneally or by the mouth. 

The yellowish kernels from six of the dry fruits were eaten by 
an adult. They produced the effect of a mild laxative. The nuts 
contained 12.96 per cent moisture and a yellow oil, representing 
28.37 per cent of the original nut. 
Table I. — Constants of the oil extracted from the dried meat of the nut of 



Quisqualis 


indica 


{ether 


extraction) 






Specific gravity at 30° 


C. 








0.9075 


Refractive index at 30° 


c. 








1.4585 


Polarization 










0.00 


Saponification No. 










187.97 


Iodine No. (Hanus) 










66.49 


Reichert Meissl. No. 










1.4 


Acetyl value 
Appearance 








clear, 


3.787 
yellow 



a Loureiro, J. Flora Cochinchinensis. Ulyssipone (1790), 1, 274. 

10 Hanbury, Daniel, Notes on Chinese Materia Medica. John E. Taylor, 
London (1862), 15. Reprinted from Pharm. Journ. for May, July, October 
(1861). 

17 Soubeiran, L. Leon, La Matiere Medical chez les Chinois. G. Masson, 
Paris (1874), 262. 



174 The Philippine Journal of Science 1917 

A slight separation of the less liquid components is observed 
when the oil has stood for several months. Ten cubic centimeters 
of the oil produced purging in an adult person without the 
hiccoughing, which is produced by taking a number of the whole 
kernels. The oil failed to show anthelminthic properties when 
given to children. 

An alcoholic extract of the residue, after petroleum ether and 
ether extractions, gives by vacuum evaporation a fragrant, pleas- 
ant-tasting substance readily soluble in water and precipitated 
from alcoholic solution by ether. When dry, it is white and 
vitreous. It strongly reduces Fehling's solution. It is probably 
a sugar. The yield is about 1.5 per cent. No apparent physio- 
logical action is noticed from the administration of it to adult 
persons or to guinea pigs. The original plant kernel after the 
above extraction was macerated with water. Much mucilaginous 
substance was obtained. This substance swells in water, forming 
a translucent solution. It is precipitated in a flocculent white 
state by alcohol. Boiled with dilute hydrochloric acid, it is con- 
verted into reducing sugars. The substance is believed to be 
vegetable mucilage or a gum, which is fairly soluble in water. 
A solution of this gum forms precipitates with most of the 
group reagents for alkaloids. It may possibly be due to this 
fact that some investigators report the presence of alkaloids in 
Quisqualis indica.™ 

After the removal of the gums from solution and evaporating 
and cooling of the filtrate, crystals of potassium sulphate formed, 
giving a yield of 3.87 per cent of purified compound. It may be 
that these two latter products are accountable for the medicinal 
properties attributed to the plant. It was not thought advisable 
with the small amount of material at hand (it being the end of 
the bearing season for sound fruit) to attempt to perform phy- 
siological tests of a conclusive nature. 

An oil was extracted from the seeds that has purgative prop- 
erties, a physiologically inactive substance resembling a sugar 
was isolated by alcohol extraction, a gum which gave many of 
the reactions of an alkaloid was extracted from the seeds by 
water, and 3.87 per cent of potassium sulphate were found. 

TYLOPHORA BREVIPES (TURCZ.) F.-VILL. (ASCLEPIdACE^E) 

Saruncad, sarungcar (II. in Zambales), pasuca (T. in Zam- 
bales), sayoncal and dail (Pang, in Pangasinan). 

" Hartwich, Carl, Neue Arzneidrogen. Julius Springer, Berlin (1897), 
282. 



xii, a. 4 Brill and Wells: Medicinal Plants, II 175 

Tylophora brevipes has been found in Zambales, Pangasinan, 
and Cagayan Provinces, Luzon, and also in Mindoro. So' far 
as we know, nothing has been written about this drug, which 
is much valued among the Filipinos of the northern provinces 
of Luzon. The Filipinos use only the fresh or dried roots in 
the form of a decoction without regard to dose or strength. It 
is prescribed in cases where a powerful and quick emetic is 
required. Herb doctors give it as a remedy against fevers, indi- 
gestion, incipient tuberculosis, asthma, and cough, to help the 
ejection of bronchial mucus. Some herbists use it as an emmen- 
agogue and even as a cure for gastralgia. It is said that if a 
patient is given a very strong dose it may produce vomiting, 
which will finally kill the patient. 

Tylophora asthmatica W. & A., a species very closely resem- 
bling T. brevipes, has considerable reputation in India for its 
medicinal qualities. Kirkpatrick is quoted as saying of his 
experience with Tylophora asthmatica:™ 

I have administered this medicine in at least a thousand cases and 
found it most valuable. In dysentery, and as a simple emetic, it is in 
every way comparable with Ipecacuanha. The dose is from 20 to 30 
grains, with half a grain or a grain of Tartar Emetic, if strong emesis 
is required. If the dysentery distinctly arises from intermittent disease, 
Quinine is conjoined. The form of the medicine I use is the powder of 
the dry leaf. If the root were used, the supply would soon be exhausted; 
besides I have found it less certain than the leaf. The preparation of 
the juice would at all times be troublesome and tedious. In catarrhal 
and chronic coughs it seems to act well. Its efficacy as a substitute for 
Ipecacuanha, not only as a simple emetic but as a remedy in dysentery, 
asthma, and catarrhal affections, is confirmed by the report of Dr. Oswald, 
Mr. Moodeen Sheriff, and others. According to the latter, the best treat- 
ment of snake bites consists in producing free emesis by the expressed 
juice of this plant, and following up its use with diffusible stimulants. 

Dymock 20 likewise speaks in glowing terms of its virtues. 
Waring 21 refers to it as country ipecacuanha. Because of the 
flattering accounts of the results arising from the use of Tylo- 
phora asthmatica and the reputation that T. brevipes already 
enjoys among the Filipinos, it was decided to make a chemical 
examination of the latter to determine if its physiological activity 

" Waring, E. J., Pharmacopoeia of India. W. H. Allen and Co., London 
(1868), 458. 

20 Dymock, W., The Vegetable Materia Medica of Western India. Educa- 
tion Society's Press, Byculla, Bombay. Trubner & Co., London. 2d ed. 
(1885), 519. 

21 Waring, E. J., Bazaar Medicines. J. & A. Churchill, London. 4th ed. 
(1883), 152. 



176 The Philippine Journal of Science im 

was due to the presence of a compound similar to the alkaloid 
found in T. asthmatica. 

CHEMICAL EXAMINATION OF TYLOPHORA BREVIPES 

Specimens of the fresh leaves and roots were used for the 
analyses. 

A petroleum ether exhaustion of the dried roots gave a slight 
amount of volatile aromatic oil and no traces of alkaloids. When 
separated from the wax and other extracted matter by steam 
distillation, the yield of yellow oil was slightly over 0.10 per cent. 
The residual plant materials were extracted with a chloroform- 
alcohol mixture. The coloring matter and resinous substances 
were precipitated with the chloroform on the dilution of the 
mixture with water acidulated with hydrochloric acid, while the 
separated alcoholic solution was concentrated to a syrup, again 
extracted with alcohol, then concentrated, and this concentrate 
extracted with water. The final orange-colored aqueous solution 
reacted positively for alkaloids. Two cubic centimeters of this 
solution given intraperitoneally to a guinea pig of 420 grams' 
weight caused death within two and one-half hours. Vacuum 
evaporation and purification through alcohol gives a small yield 
of the alkaloid in the form of acicular crystals often grouped 
and radiating (Plate I, fig. 1). Administered to guinea pigs, 
0.10 gram of the colorless crystals produced spasms and symp- 
toms of vomiting. An acid water extract of part of the original 
ground plant material- clarified by the neutral lead acetate method 
and concentrated gave identical crystals from an alcoholic ex- 
traction of the concentrate. Concentrated sulphuric acid gives 
a red color, while nitric acid gives a magenta red color with this 
compound. The crystals have a peculiar, somewhat acrid 22 and 
bitter taste and are soluble in water and in alcohol. The quantity 
separated was insufficient for quantitative chemical analysis. 

In 1890 Hooper, working on Tylophora asthmatica, isolated an 
alkaloid which he termed "tylophorine." The presence of such 
a compound in the Asclepiadacese has been noticed by several 
investigators, notably by Broughton, 23 of Ootacamund, in 1872, 
when he obtained a small quantity of crystals insufficient for 
analysis. Extracts from the leaves of Tylophora brevipes con- 
tain an alkaloid reacting like that found in the bark. 

The work done by us on Tylophora brevipes shows the presence 

" Fluckiger, F. A., and Hanbury, D., Pharmacographia. A History of 
the Principal Drugs of Vegetable Origin. Macmillan & Co., London. 2d. ed. 
(1879), 427. 

23 Fluckiger and Hanbury, loc. cit. 



xii. a. 4 Brill and Wells:' Medicinal Plants, II 177 

> of an alkaloid which is similar in action to tylophorine, found by 
Hooper -' in Tylophora asthmatica W. & A. and recognized in 
the dispensatory. 25 

The local plant can be used in the same manner as T. asthma- 
tica, since its action is due to the presence of an alkaloid iden- 
tical with or closely resembling tylophorine. 

TODDALIA ASIATICA (L.) KURZ (RUTACE^E) 

Dauag, dauag manoc, cayutanang baguing (T. in Rizal), 
atangen and bugkan (Ifg. in Benguet). 

This plant is found in many places of the Philippine Islands, 
being a species of wide geographic distribution. 

Notwithstanding the fact that the root of this plant has been 
widely known for several centuries for its febrifugal and anti- 
diarrhoeal properties, even to the extent of having been added 
to some of the European pharmacopoeias under the name of Radix 
indica lopeziana, or "root of Juan Lopez Pigneiro," Filipinos 
know nothing of its medicinal properties. 

Dymock, 26 says: 

This scandent shrub appears to have been one of the plants known to 
Sanskrit writers as kanchana or golden, on account of the orange color 
of its fruit. 

Yet, in his compilation of the Sanskrit books about medicine, 
Udoy Chand Dutt says nothing about this drug, although its use 
is apparently general in India. 

Fliickiger and Hanbury 27 say that the root of Toddalia was 
first known in 1671, thanks to Doctor Redi, an Italian, who de- 
scribed it from the samples in his possession gathered by Juan 
Lopez Pigneiro near the mouth of Zambesi River, East Africa. 
In 1771 this drug was introduced into European medicine by 
Gaubius as a remedy against diarrhoea. Its reputation became 
so great that in 1792 it was included in the Pharmacopoeia of 
Edinburg, although its botanical origin was not noted. As often 
happens with vegetable drugs without any specific action, Radix 
indica lopeziana has fallen into disuse. 

The chemical composition of Toddalia root is not yet clear, 
notwithstanding the analytical researches done by Fliickiger and 
Hanbury and also by Dymock. The former say that they were 

"Hooper, David, Pharm. Journ. & Trans. (1891), 21, 617. 

- 5 Hare, Caspari, Rushy, National Standard Dispensatory. Lea & Febiger, 
Philadelphia (1908), 269. 

w Dymock, W., Pharmacographia Indica. Kegan Paul, Trench, Triibner 
& Co., Ltd., London (1893), 1, 260. 

27 Op. cit., 777. 



178 The Philippine Journal of Science 19" 

unable to prove the presence of berberine 28 in the drug, and the 
latter studied only the volatile oil obtained by the distillation of 
the leaves. Nadkarni -'■' is of the opinion that the root contains 
berberine. 

In view of this difference of opinion George Watt, author of 
the Dictionary of the Economic Products of India, bewails the 
fact that none of the Hindu chemists has endeavored to find 
out the composition of the fresh material. 

Some laboratory work has been undertaken by us with the 
object in view of determining whether the intensely bitter taste 
of the root of Toddalia asiatica is due to berberine, to some bit- 
ter principle, or to an alkaloid of an unknown nature. 

Following is the special opinion of Moodeen Sheriff regard- 
ing the virtues of Toddalia asiatica, as it is given in Watt's 
dictionary : 

I have been using the root-bark in my practice for the last sixteen or 
seventeen years, and do not hesitate to say that it is, as an antiperiodic 
and antipyretic, equal if not superior to quinine and other alkaloids of 
Cinchona and to Warburg's tincture respectively. As a diaphoretic, it is 
decidedly superior to Pulv. Jacobi Vera, and as a tonic to Gnetian and 
Calumba. It is highly useful in effecting a cure in all idiopathic and 
uncomplicated fevers, whether periodical or continued. It is best used in 
tincture and decoction and I make these preparations three or four times 
stronger than those generally in use. This is the chief reason, I think, which 
has rendered the drug so successful in my hands. The analogy between 
the medicinal properties of the root-bark of T. aculeata and those of the 
root of Berberis aristata and a few other species of Berberis is very great 
and complete, there being no difference whatever. Therefore, everything 
I have said about the preparations, doses, therapeutic used and the manner 
of using the tincture and decoction of the latter is quite applicable to those 
of the former. The drug under consideration, however, has one great 
advantage over the root of Berberis aristata and other species of Berberis, 
namely, that it is procurable in every large bazar of Southern India; where- 
as the roots of the latter plants must be procured from distant places, 
such as the Nilghiris, Shevaroy Hills, Central and Northern India, etc., 

The following formulas belong to the Pharmacopoeia of India 
(1868) : 

Tincture of Toddalia (Tinctura Toddalia). Take of the root-bark of 
Toddalia, bruised, two ounces and a half; Proof Spirit, one pint. Macerate 
for seven days in a closed vessel, with occasional agitation; strain, press, 
filter, and add sufficient proof spirit to make one pint. It may also be 
prepared in the same manner as Tincture of Calumba. 

Dose: From one to two fluid ounces twice or thrice daily. 

28 Op. cit. 

" 9 Nadkarni, K. M., Indian Plants & Drugs with their Medical Properties 
and Uses. Norton and Co., Madras (1910), 398. 



xii. a. 4 Brill and Wells: Medicinal Plants, II 179 

Infusion of Toddalia (Infusum Toddalia). Take of the root-bark of 
Toddalia, in coarse powder one ounce; boiling water, ten fluid ounces. 
Infuse in a covered vessel for one hour, and strain. 

Dose: From one to two fluid ounces twice or thrice daily. 

CHEMICAL EXAMINATION OF TODDALIA ASIATICA 

The brittle root bark cleared of the outer layer was finely 
ground, dried at a low temperature, and exhausted in the usual 
way with petroleum ether, ether, alcohol, and water. 

Petroleum ether removed a small quantity of wax and a little 
resin, which was dissolved in a sharp-tasting pungent oil. It 
is possibly due to the not unpleasant taste of this oil that parts 
of the plant are used in India as a condiment. The resin is 
easily crystallized from an ether-alcohol mixture in groups of 
elongated prisms or needles. 

An absolute alcoholic extraction of the dried plant residue 
gives a yellow solution which, when purified by repeated evapo- 
rations and alternate extractions with water and alcohol, yields 
a final aqueous solution of a greenish yellow color, giving the 
group reactions of alkaloids. A few of the reactions are as 
follows : 

Gold chloride, yellowish precipitate; platinic chloride, no pre- 
cipitate; Mayer's solution, white precipitate; phosphomolybdic 
acid, white precipitate; potassium cadmium iodide, white pre- 
cipitate; mercuric chloride, white precipitate (light) ; potassium 
mercuric iodide, white precipitate; Kraut's reagent, ochraceous 
orange precipitate; picric acid, yellow precipitate; bromine, 
orange precipitate (heavy) (darkens). Vacuum evaporation 
leaves a thick amorphous alkaloid. A portion diluted with al- 
cohol, spread in thin sheets and allowed to evaporate slowly, 
formed crystals (Plate I, fig. 2). Both the hydrochloride and 
sulphate were made, and a small quantity of crystals of each 
was obtained. The free alkaloid both in the amorphous and 
crystalline condition is soluble in the cold in distilled water, 
giving a bright yellow color, and is fairly soluble in methyl 
alcohol and ethyl alcohol, slightly soluble in acetone, chloroform, 
and ethyl acetate, while only slight traces dissolve in ether, petro- 
leum ether, benzene, and amyl alcohol. From this may be under- 
stood the property which it possesses of not being easily sep- 
arated from either alkaline or acid solutions by agitation with 
immiscible solvents. 

The alkaloid has a persistently bitter taste, but is not as in- 
tensely bitter as the alkaloid found in Lunasia amara Blanco. 
The crystalline forms of the free alkaloid and of its salts seem 
identical with those of berberine, with which comparative tests 



180 The Philippine Journal of Science i»m 

were made. The alkaloid conforms to all the tests given by 
Mulliken 30 for berberine. Like berberine the salts have an in- 
tensely yellow color and an extremely bitter taste. A solution 
of iodine in potassium iodide gives a brown precipitate of B.HLI 2 , 
crystallizing from hot alcohol, in long adamantine needles, which 
are insoluble in water and cold alcohol. Warming with concen- 
trated sulphuric acid gives an olive green color. A fragment 
of NaNO;,, stirred into the solution in H 2 S0 o gives a violet streak. 
Chlorine water gives the characteristic purple zone effect. By 
the careful addition of weak nitric acid acicular crystals of 
the yellow nitrate are formed (Plate II, figs. 1 and 2) . The salts 
are decomposed by heat. These reactions are characteristic of 
berberine and of the alkaloid from Toddalia asiatica. Since 
the alkaloid responds to all the characteristic reactions for ber- 
berine, we feel justified in concluding that Toddalia asiatica 
Kurz of the Philippine Islands contains berberine. 

LUNASIA AMABA BLANCO (RUTACE^E) 

Lunas (T. in Rizal, Bataan, Bulacan, Mindoro, and Pam- 
panga), paitan, pait, lunas bondoc (T. in Bataan), malasanqui, 
lunas na puti (T. in Rizal), bayabayabasan, saltiqui, santiqui 
(T. in Laguna), abdong cahoy (T. in Batangas and Laguna), 
malaligas na babae (T. in Baler), malacacao, cacaocaocaoan (T. 
in Batangas), lubilubi (V. in Cebu), labao (V. in Ticao Island), 
paetan (V. in Leyte and other southern Islands), pait-pait (V. 
in Zamboanga), and saguit (M. in Zamboanga). 

Lunasia amara is a species widely scattered throughout the 
Archipelago; it is found in most provinces and islands. 

The literature of Lunasia is scant, for up to the time when 
Dr. W. G. Boorsma :ri began his investigations with authentic 
material of this species all that had been published with refer- 
ence to its chemical composition was in reality about the bark of 
abuab (Lophopetalum toxicum Loher) , which is utilized by some 
of the inhabitants of the Philippines to poison their weapons. 

This unfortunate confusion arose from the fact that Lophope- 
talum toxicum had been sent to European museums by Dr. Scha- 
denberg, a resident of Manila at the time, with the erroneous 
name of Rabelaisia (Lunasia) . Plugge, and other investigators, 
who could only avail themselves of the material received at the 
museums, were victims of this error. 

s0 Mulliken, S. P., The Identification of Pure Organic Compounds. John 
Wiley & Sons, London (1916), 2, 261. 

"Boorsma, W. G., Bull. Inst, Bot. Buitenzorg (1900) 6, 15. 



xii. a. 4 Brill and Wells: Medicinal Plants, II Igl 

Thus rebelaisine, of Plugge, should be therefore called lophope- 
taline, a substance utterly different from lunasin, which was 
obtained by Lewin and later identified with that obtained by 
Boorsma from Lunasia costulata Miq. (=L. amara Blanco). 
This author, in his commentary on the results of his research, 
says that he obtained an amorphous alkaloid, hygroscopic, of 
bitter taste and nonvolatile, from the bark of Lunasia costulata. 
Lunasin was dark grayish and was very soluble in water. Eight 
milligrams in a hypodermic injection acted as a cardiac poison 
and killed a frog of 78 grams' weight; the heart stopped in 
systole. A dose of 20 milligrams injected into a guinea pig 
produced a severe intoxication with complete paralysis for some 
time, though without occasioning death. Lunasin has been also 
found in the hard wood of the tree. 

Father Juan J. Delgado 32 mentions three species of paetan — 
the first and second identified by Fernandez- Villar as Lunasia 
amara Blanco. There are several distinct species of Lunasia 
found in the Philippines, so that perhaps Delgado's statements 
really refer to more than one form. 

Following is a translation of a part of what Delgado writes 
concerning lunas or paetan. 

Paetan is a name common to three species, all of them very bitter, from 
which it derives its name, which we might translate as "bitterness itself." 
This is true of the one that is larger than the rest, whose branches grow 
to be of the size of the wrist. I wished to make experiments on this one, 
and for that purpose had some brought from Panamao Island. It is very 
bitter and acts as a poison. By merely chewing a very small bit of its 
bark, I was hardly able that day to remove the bitterness from the tongue, 
experiencing some sort of convulsions for three days. I was obliged to take 
some antidotes and cold water baths, but after the three days were over, 
I felt exceedingly robust and well. 

All this leads me to believe that if this plant were carefully studied, 
its value as a medicine would be learned. A lad who tasted it at the same 
time I did immediately felt the same effects, but having in his case pro- 
voked nausea, was soon well. Some of the natives use an infusion of the 
iMrk of paetan to cure eye trouble; the solution is so strong that the 
bitter taste becomes evident in the throat and mouth. 

The second paetan is a smaller tree, which is found in the smaller 
islands of the Archipelago, and is very well known and much valued for 
its medicinal fruits. It is an antidote for all poisons whether they be from 
witch-craft or from snakes, or poisonous animals or fish. Half a real in 
powdered form will cure a severe stomach ache due to any cause ; it is 
admirable for infected wounds, for it will clean and cure them in a short 
time. Finally it is one the antidotes used to make the "oleo de japlas." 

32 Delgado, J. J., Historia general sacro-profana, politica y natural de las 
islas del Poniente llamadas Filipinas (1892), 611. 



182 The Philippine Journal of Science 1917 

I honestly believe that it would be a fine remedy for sciatica and gout; 
those who suffer from these diseases may try without any fear. 

From the above it appears that instead of being a harmless 
drug lunasine should have very strong properties. Attention 
is called to the fact that the author does not mention the use 
which, according to some, the Filipinos made of it to poison 
arrows. This statement, so often made by the foregoing writers, 
is perhaps a misinterpretation of Blanco's statement in the 
original description of the species. A translation of his remarks 
regarding the use of the wood follows: 

The grain is very fine and close, and the wood is so hard that the 
Negritos use it instead of iron for the points of their arrows. 

We are almost convinced of this misinterpretation, for not- 
withstanding our careful investigations we. have not been able 
to find a Filipino who could corroborate the assertion made by 
those who have written about Lunasia being used to poison 
arrows. 

The investigation of medicinal and poisonous plants which the 
Bureau of Science is carrying on at the present time has no fur- 
ther available information in regard to the immediate medicinal 
uses of the species here in question than that obtained from 
Laguna Province, which translated is as follows : 

The juice of either the dried or fresh seeds, macerated in alcohol or 
cooked in oil and then concentrated, is generally used against the bite of 
poisonous animals. 

It is the general opinion among herb doctors that not only 
the bark, but also the seeds are a good remedy to cure gastralgia 
in general and certain adynamic conditions of the digestive 
organ. 

CHEMICAL EXAMINATION OF LUNASIA AMARA 

The bark and leaves were used for analysis. 

From the prepared bark, petroleum ether extracts a small 
quantity of vegetable wax and traces of alkaloids ; ether extracts 
contain no traces of alkaloids, but a small quantity of wax, not 
removed by the petroleum ether, and the chlorophyll existing 
in the bark. Following the ether, extraction a mixture of equal 
parts of ethyl and methyl alcohols removes a considerable quan- 
tity of alkaloidal substance. 

Purification by repeated alternate extractions with water and 
alcohol leaves the alkaloid in an amorphous condition. The yield 
is about 0.6 per cent. It is extremely bitter, has a brownish 
color, and is yellow in alcoholic solution. It is readily soluble 
in water and takes up water when exposed to the air. It is 



xii. a. 4 Briil and Wells: Medicinal Plants, II 183 

soluble in hot absolute ethyl and methyl alcohol, from which 
solutions by careful evaporation crystals were obtained (Plate 
III). 

The amorphous compound is but slightly soluble at 20° C. in 
ethyl and methyl alcohol. It is almost insoluble in cold chloro- 
form, ethyl acetate, and petroleum ether and is insoluble in cold 
ether, benzol, amyl alcohol, and acetone. The amorphous sub- 
stance precipitated from the ethyl-methyl alcoholic solution by 
ether is very hygroscopic, quickly changing from a white solid 
to a reddish brown aqueous solution. Plate III, fig. 1, shows the 
dendritic crystals which formed quickly from absolute alcohol. 
Plate III, fig. 2, shows cubical crystals formed from the amor- 
phous substance on long standing in a vacuum desiccator over 
calcium chloride. Both forms of crystals were soluble in the 
organic solvents that dissolve the amorphous alkaloid and were 
toxic in large dosage in their action toward guinea pigs. 

The hydrochloride, nitrate, and sulphate were made. The 
hydrochloride crystallizes with difficulty in acicular form from 
a weakly acid solution (Plate IV, fig. 1). 

Besides the above colorless crystals a few prisms may be 
observed present in the microphotograph. Both forms of crys- 
tals react positively for alkaloids. This suggests possibly the 
presence of two alkaloids. 

The nitrate forms in pale yellow elongated prisms, and the 
sulphate forms in plates or prisms of the same color. An 
aqueous solution of the purified free alkaloid reacts heavily with 
the general group reagents: 

Iron chloride gives a reddish solution ; gold chloride, a yellow 
precipitate ; platinic chloride, a pale yellow precipitate ; bromine, 
a yellow precipitate; potassium-cadmium-iodide, a white precip- 
itate; mercuric chloride, a white precipitate; potassium iodide, 
a white precipitate (slight); picric acid, a yellow precipitate; 
Mayer's reagent, a white precipitate; Kraut's reagent, a red 
precipitate. 

The compound was tested for nitrogen; it reacted positively. 
The bromine compound is unstable, easily decomposing when 
filtered free of excess bromine. A portion of the crude alkaloid 
was purified by precipitation with Kraut's reagent and broken 
up by alkaline carbonate. The pure alkaloid obtained was not 
a large yield and was amorphous. 

The dried, finely ground leaves were exhausted in the usual 
manner with petroleum ether, ether, and chloroform-alcohol mix- 
ture. The petroleum ether solution contained a neutral wax 
and a small quantity of an essential oil of yellow color and sharp 



184 The Philippine Journal of Science isn 

taste and no traces of alkaloids. The ether solution gave a small 
quantity of aromatic acid resin and traces of the same essential 
oil. After dilution and precipitation of the coloring matter with 
the chloroform, the aqueous solution was evaporated on the steam 
bath, and the product was purified according to the method 
previously used. The same alkaloid was found to be present 
in the leaves, but seemingly in lesser quantity than in the bark. 

In both cases the exhausted plant material was finally macer- 
ated with hot water. The water extract showed but little of the 
alkaloid not extracted by the organic solvents. Neither acid 
nor alkaline extractions of the original plant material were made 
for the reason that evaporation of such solutions prepared from 
the isolated alkaloid showed a marked decomposition on heating. 
Solutions of the free alkaloid as prepared from approximately 
neutral solution could be evaporated to dryness without vacuum 
and without any apparent decomposition. The salts were readily 
decomposed by warming on the steam bath. 

Two cubic centimeters of a 10 per cent solution of the free 
alkaloid produced apparent, great weakness when injected into 
a guinea pig. The animal recovered. About 0.8 gram given to a 
dog produced apparent weakness, a desire to sleep, and a slow- 
ing of the heart action. There were no signs of pain. The 
animal recovered its normal activities after two hours had 
elapsed. 

Lunasia amara contains an alkaloid or possibly two alkaloids 
that have some physiological activity. 

ROUREA ERECTA (BLANCO). MERRILL (CONNARACE^) 

Palo santo (Sp.-F. in Rizal), camagsang baguing (T. in Rizal), 
lenamo (Cuyo Island), malagranada (F. in Bulacan), and gango 
(II. in Ilocos Norte). 

The name camagsang baguing is better applied to Rourea 
volubilis (Blanco) Merrill. 

The family Connaraceae, to which this plant belongs, yields 
few drug-producing plants. With the exception of two species 
of Agelaea, two of Connarus, one of Cnestis, and another of 
Rourea, all are employed in popular medicine wherever they 
grow, though none go beyond the limits of quackery. The 
family is of slight importance in materia medica. The seeds of 
Rourea are poisonous. 

It may be generally stated that the Connaracese are charac- 
terized by the presence in their tissues of astringent elements 
and balsamic resins. Judging from the uses indicated by Fili- 



xii, a, 4 Brill and Wells: Medicinal Plants, II 185 

pinos, the roots of Rourea erecta are more astringent than 
balsamic. 

The data gathered up to date in regard to the use of palo 
santo in the Philippine Islands are to a certain extent contra- 
dictory. Some say that a decoction from the roots up to the dose 
of a teaspoonful is an emetic, but if this dose is exceeded, it is a 
poison. This decoction mixed with the food and given to hogs 
and dogs will kill them. The animal becomes nauseated and 
dies, but if nausea does not follow, the animal will swoon and 
will expire without recovering. 

Information received from La Union Province presents similar 
evidence with some details of the symptoms as follows : 

The animal staggers and falls to the ground moaning and expires after 
a while. 

Others claim that death does not follow immediately, but after 
an interval of three or four days. 

Other information received from Samal (Bataan) is that the 
decoction from the bark of the roots expedites childbirth and 
that the fruits are poisonous and are used to kill dogs. 

And lastly, reports from Pangasinan do not mention the roots 
or the stems, but state that the fresh or dried leaves in decoction 
are used to cure gastralgia and are an absorbent. 

The medicinal properties which Delgado 33 attributes to palo 
santo (guicos guicos, hanmabao) , and according to Blanco also 
known as camagsd taquilis (Tag.), ungali na mapuld magtabig, 
and mavindato (Visayan and Pampango), are those that cause 
the natives from the vicinity of Manila to make use of the roots 
of the species herein mentioned. However, guicos guicos is not 
Rourea erecta (Blanco) Merrill, but it. volubilis (Blanco) Merrill, 
which was identified by Fernandez-Villar as Rourea heterophylla 
Planchon, a synonym of R. volubilis Merr. 

Delgado states that one of the medicinal vines for many ail- 
ments is that which is called by the natives of some islands 
guicos guicos, while in others they call it hanmabao. The 
Spaniards call it palo santo, due to its admirable properties. It 
is a fine remedy for lockjaw and colds, which are common mala- 
dies in this country. The decoction of this vine is very good 
for those who suffer from venereal diseases, and has sudorific 
properties. 

De Mercado, in his Libro de Medicinas de esta Tierra, and 
Blanco, in his Flora de Filipinas, also speak of palo santo, which 

83 Delgado, J. J., Historia general sacro-profana, politica y natural de las 
islas del Poniente llamadas Filipinas (1892), 780. 



186 The Philippine Journal of Science iei7 

is called by the former camunin or guayacan, as a sudorific and 
purgative. 

m Having in mind the opinion of these three authors about 
guicos guicos and the scant information about camagsa, which 
is more commonly known as palo santo, it is easily understood 
how the Filipinos of to-day have mistaken the two species — con- 
fusing R. erecta and R. volubilis and making use of the former 
instead of the latter. It is possible that the three observers 
were misinformed about the botanical origin of the medicinal 
species or it is possible that both plants have the same properties, 
though the latter supposition seems untenable. 

CHEMICAL EXAMINATION OF ROUREA ERECTA 

The roots and fruit are the parts used. 

Extractions from the roots show a high percentage of plant 
wax, no alkaloids, glucosidal bodies, or any active principles. 
After the extraction with ether, extraction with ethyl alcohol (92 
per cent) gave a thick mass that amounted to 24 per cent of the 
original plants when evaporated under 600 millimeters vacuum. 
This mass consisted of tannin substances, not alkaloidal in 
character (phlobaphene), and other coloring matter. The color- 
ing matter is a bright magenta. It is red in acid solution and 
gives a dark green precipitate in alkaline solution. The red color 
is soluble in water and easily washed free from the other solid 
extracted matter. It is also easily decomposed by concentrated 
hydrochloric acid. 

Water extractions of the roots gave no decided symptoms when 
administered to a dog or injected into a guinea pig. 

The berries were analyzed both in the fresh and dried state. 
The toxicity of the ground, fresh berries was not greater than 
that of old specimens. In fact, the highest toxic effect was 
obtained from a residual worm-eaten specimen that had had 
all opportunity to ferment and undergo decomposition. Robert 34 
has made the statement that the active principle of Rourea 
oblongifolia is destroyed when the plant dries. This is not true 
in the case of R. erecta. - Experiments were made on healthy dogs 
using the ground berries and the numerous products of extrac- 
tion. Taking the freshly picked, finely ground berry and admin- 
istering it in meat to a dog, no effect was noticeable until the 
third day, when slight weakness was apparent. These symptoms 
were manifested after the consumption of 20 grams of the 
berry in a period of seventy-two hours. Only slight, increasing 

31 Robert, R., Centralbl. f. klin. Med. (1893), 14, 930. 



xii. a, -i Brill and Wells: Medicinal Plants, II 187 

weakness was noticeable up to the sixth day. During the fol- 
lowing thirty-six hours the dog lost control of voluntary move- 
ment, and the activity of the salivary glands seemed greatly 
increased, a large volume of clear liquid passing from the mouth. 
The pupils were dilated, but vision seemed good. There were 
no convulsions before death. The dried fruit gave similar results 
when administered in much smaller dosage. Five grams of dry, 
worm-eaten seeds were finely ground and given to a healthy dog 
of full growth. Death followed after a period of thirty-eight 
hours. An examination showed no abnormal conditions. There 
was no congestion, the bladder was empty, and the liver, spleen, 
kidneys, and lungs were normal. The heart was full. Renson 35 
states that carnivorous animals are affected by eating the berries, 
while birds and fowls consume it with impunity. 

The dried seeds, finely ground and extracted with petroleum 
ether, gave a white crystalline mass containing a small amount 
of ether-soluble resin of a yellow color. Attempts were made 
to oxidize this white product from the berries (melting point 
64° C.) by heating with potassium permanganate in sodium car- 
bonate solution for two hours on a steam bath; the manganese 
oxides decomposed by sulphurous acid, and the resulting product 
recovered. The compound had the melting point of the original 
body (64° C). No oxidation took place under these conditions. 

The compound was treated with ammonium hydroxide in 50 
per cent alcohol solution and sodium carbonate in excess on the 
water bath for two hours ; hydrochloric acid was added, and the 
mixture was filtered. The product melted at 64° C, the same as 
the original sample. Also the original compound was treated in 
acetic acid with bromine, by heating on the steam bath for one 
and one-half hours. When the solution cooled, a flocculent preci- 
pitate resulted. This precipitate, washed free of acetic acid 
and bromine, gave a melting point of 63° C. ; the orignial com- 
pound was recovered. The compound showed no physiological 
effect, and it was concluded to be plant wax. The plant residue 
left after the extraction by petroleum ether was extracted with 
cold and hot absolute alcohol. No toxic substances were found 
in this solution. Precipitation was obtained from the alcohol 
extract by dilutions with water. An aliquot part of this precipi- 
tate was vacuum dried and extracted by solvent. The products 
•were administered in large dosage to dogs without resulting 
effect. Extracts were obtained from this same plant material 
using consecutively the solvents ether, chloroform, amyl alcohol, 

"Renson, Carlos, Pkarm. Journ. & Trans. (1891-92), 22, 982. 



188 The Philippine Journal of Science uh 

and ethyl acetate. When thought necessary, these extractions 
were on material made acid or alkaline. The extracts freed from 
the solvent produced no physiological effects on dogs or guinea 
pigs. In no case was any disturbance in the activity of guinea 
pigs observed from the administration of any portion of this 
plant. These results confirm Renson's conclusions regarding the 
nonpoisonous character of the plant toward the Herbivora. 

After the extractions with organic solvents neutral warm 
water was used to extract the plant. Vacuum evaporation of 
the filtered solution gave a small mass of crystalline substance 
which was recrystallized once from methyl alcohol followed by 
ethyl alcohol. The quantity was insufficient for analytical pur- 
poses. Eight milligrams were given a healthy 1-year-old dog. 
The compound exhibited the same effect as that produced by 
the crude drug. In this case the animal showed a change in 
mental condition similar to that observed by Renson 88 in his 
"second form of poisoning." Although the plant material con- 
goura (Rourea oblongifolia) used by Renson is not the same as 
that used by us, yet our results indicate that these plants contain 
the same active principle. He describes what he terms the 
"second form of poisoning" in the above reference as follows : 
First or Acute Form of Poisoning, followed by Death. 

Two or three hours after the hypodermic injection the animal commences 
to vomit without great effort, it expels fecal matters, and trembles a little. 
An increase of activity in the salivary glands is than observed, a clear and 
limpid liquid trickling from the mouth, drop by drop and in quantity rela- 
tively great, as though jaborandi had been taken. Soon the movements in 
walking become uncertain, the animal stumbles, and appears intoxicated. 
Later it remains lying down and there is a loss of voluntary movements. At 
last convulsions occur, the mouth remains open, and the pupils are dilated. 
These convulsions are accompanied by moans and blinking of the eyelids. 
Then appears a most curious phenomenon. The dog, although lying on its 
side, executes the movements of walking. With ears and tail erect, it shakes 
its head in an almost lively manner, and there is no doubt it believes it 
walks. This dream, if I may so call it, immediately follows the convulsive 
attack. The attacks are repeated a great number of times, and in the 
intervals there are hiccups and sudden starts. At last the animal, com- 
pletely worn out, falls into a comatose state, its respiration becomes un- 
equal, it emits a little blood by the nostrils (in certain cases, but not in 
others), and dies at the end of ten to twelve hours. 

Second Form of Poisoning, followed by Recovery. 

It suffices to introduce half a cubic centimeter of the liquid extract of 
Cangoura into the organism, either by means of the skin or the stomach, to 
produce the following effects in a dog of average size. The day of injection 

" Loc. cit. 



xii, a. 4 Brill and Wells: Medicinal Plants, II 189 

and the following day the animal appears perfectly well. Two days after- 
wards the cerebral disturbances appear. Suddenly, without apparent cause, 
it rushes about in an unbridled career, barking loudly, and seeming to pur- 
sue an imaginary prey. After that it stops, with its tail between its legs, 
and begins to howl in a very disagreeable fashion. It now stumbles left and 
right, so that one might believed it drunk with alcohol. During the fit it 
scarcely lets itself be approached by its master, whom it regards with a 
threatening air. Other attacks then occur, during which it involuntarily 
expels solid and liquid excrementitious matters, whilst it indulges in a furious 
and noisy run. Suddenly it stops, seized with a slight nervous trembling, 
and as it shivers one would say it had cold. The strange way in which it 
then begins to sniff the air causes the belief that there is also a disordered 
sense of smell. Now, abruptly springing up, seized with a great terror, and 
with its tail between its legs, it runs away crying, and hastens to take refuge 
in dark corners or under the furniture, or it seems to defend itself by bark- 
ing at an imaginary phantom. It is these dreadful visions which above 
all predominate, but they soon give way to attacks of rage, or to a stupid 
desire to make a noise. The fear is very great and the sensibility over- 
excited, and any object, however slight, such as a small piece of wood, 
thrown to the dog during the fit causes it to utter loud cries. When it be- 
gins to drink the reflections from the water as it is agitated appear to make 
the animal afraid and it shrinks away. Later it is seized with an attack of 
mastication, and flakes of white foam issue from its mouth. The salivary 
secretion continues very excessive, though intermittently, during three days, 
in the course of which the dog is observed to slumber momentarily, with 
open eyes; then abruptly, at the end of some seconds, it moves as though 
awaking; in brief, it has lapses of consciousness. This is very characteristic. 
All these symptoms are repeated during six to eight days, after which the 
dog is restored to health. Subsequently it remains for some time subject to 
epileptic-like fits, but the animal, which had become of a gloomy and sad 
disposition, recovers little by little its normal state. 

The condition produced in the dog by the 8 milligrams of 
substance lasted a period of six days. On the seventh day food 
was taken and the animal had fair control of his muscles. On 
the tenth day he seemed to have fully recovered his previous con- 
dition, except for weight. 

Robert 3T also speaks of the remarkable effect of the crude 
drug on dogs and states that it might contain a compound of 
therapeutic value. With the beginning of the next fruiting sea- 
son, when larger quantities will be available, the investigation 
of this plant will be continued. 

Without doubt Rourea erecta is poisonous in small doses to 
carnivorous animals. Rourea volubilis may not be poisonous, 
or the active substance may not be present in so large quantities. 
An attempt will be made to obtain specimens of the latter for 
examination in order that its properties may be compared with 
those of R. erecta. 

" Kobert, R., Centralbl. f. klin. Med. (1893), 14, 927. 



190 The Philippine Journal of Science iw? 

HYMENODICTYON EXCELSUM WALL (?) 88 (RUBIACE.E) 

Tubo-bato (Tagbanua in Palawan), abar (II. in Abra and 
Uocos Sur), alegango (T. in Bulacan), mag-talisay (V. in Gui- 
maras Island), balangcari (T. in Nueva Ecija), malatabaco 
hibao (T. in Rizal), and huliganga (T. in Batangas). 

This rubiaceous plant, which once acquired some renown in 
India, Blanco erroneously classified as Exostemma philippicum. 

In view of the opinion of some writers who claim the possibil- 
ity of using its bark as a substitute for quinine as a febrifuge, 
a chemical trial was made to obtain the hymenodictyonine, the 
alkaloid obtained by Naylor :9 from the Indian Hymenodictyon 
excelsum. We were unable to isolate the alkaloid described 
by Naylor. 

The fact that the alkaloid that apparently characterizes the 
Indian Hymenodictyon excelsum Wall, was not found in the bark 
of Philippine Hymenodictyon raises some doubt as to whether or 
not the plant found in the Philippines belongs to that species. 
Our identification of the Philippine material as H. excelsum 
follows the determination of Philippine material made at Kew, 
England. 

Following we quote what Dymock writes in his vegetable 
materia medica of Western India : 40 

In 1870, Broughton examined the fresh bark of one of the Hymenodic- 
tyons, and found that the bitter taste was due to the existence of aesculin, 
and that the bark when dry was almost tasteless owing to the transforma- 
tion of that substance into aesculetin the decomposition having been in- 
duced by contact with decaying organic matter. 

The fact here mentioned that the bark when dry lost its bit- 
terness leads us to suppose that it was not that of H. excelsum 
but of H. obovatum, the dry bark of the former tree being ex- 
tremely bitter. 

CHEMICAL EXAMINATION OF HYMENODICTYON EXCELSUM 

Both the fresh and dried bark were used in these analy- 
ses. The fresh bark treated with water and heated to 90° C. 
gives a highly satisfactory extraction of the active principle. 
The fluorescent filtrate was clarified with neutral lead acetate 

3S The Philippine plant is a Hymenodictyon very closely allied to, if not 
identical with, H. excelsum Wall. A critical comparison with type material 
will be necessary definitely to determine whether one or two species are 
involved. 

33 Naylor, W. A. H., Pharm. Journ. (1883-4), 14, 311. 

40 Dymock, W., The Vegetable Materia Medica of Western India. 2d. ed. 
(1885), 404. 



xii. a. 4 Brill and Wells: Medicinal Plants, II 191 

according to the method of van Rijn " and was evaporated. A 
crystalline substance was obtained. This compound reduced 
Fehling solution on prolonged boiling. Tests for alkaloids by 
the usual methods showed no such compounds present. 

Four hundred grams of fresh bark were macerated with hot 
water, the extract was filtered, and the filtrate was evaporated 
to a thick syrup under vacuum and then repeatedly extracted 
with 80 per cent ethyl alcohol. A yield of 11.091 grams of 
pale yellow crystals was obtained. When purified by recrystal- 
lization from hot dilute alcohol, this compound forms a mass 
of colorless crystals by quick crystallization and groups of plates 
or flat prisms by slow crystallization. 

Four hundred grams of quickly dried bark were first extracted 
by petroleum ether and then by ether. Neither of these solu- 
tions contains alkaloids and but slight traces of the above- 
mentioned compound. The residual plant material was treated 
with a mixture of 90 per cent alcohol and chloroform and filtered 
hot. This solution treated with water liberates a considerable 
amount of crystallizable compound of the same nature as that 
obtained by the method of van Rijn. It is purified by repeated 
crystallization from dilute alcohol. 

The nature of the active principle of Hymenodictyon excelsum 
seems to be accepted by some as sesculin and by others as an 
alkaloid. Broughton 42 found sesculin in Hymenodictyon excel- 
sum by extracting the bast material with hot water, evaporating 
the filtered solution, mixing it with magnesia, and evaporating 
and extracting the residual mass with strong alcohol. By con- 
centrating this filtrate and letting it stand twelve hours, he 
obtained prismatic crystals which on acid hydrolysis gave sescu- 
letin and a sugar. Broughton also found that the bitter taste is 
not obtained from the dry bark, and he concludes that in drying 
the sesculin is converted to sesculetin. 

Naylor 43 denies Broughton's conclusion of the compound being 
sesculin and alleges that Hymenodictyon excelsum contains an 
alkaloid which he terms "hymenodictyonine." Naylor's second 
method of extraction and the one he recommends is as follows : 

Mix the ground bark with I weight of lime and convert into a thick 
paste with H 2 0, dry and exhaust the mixture with chloroform. Extract 
the chloroform mixture with dilute sulphuric acid solution, precipitate the 
alkaloid with sodium hydroxide solution, and extract with ether. Repeat 
to purify. 

41 Op. cit., 288. 

12 Broughton, Jahresber. d. Pharmacog., Pharm., u. Toxic. (1868), 3, 81. 

"Naylor, W. A. H., Pharm. Journ. & Trans. (1883-4), 14, 311. 



192 The Philippine Journal of Science isn 

He prepared a crystalline hymenodictyonine by extremely slow 
evaporation of an ethereal solution. Analysis of the compound 
showed it to be dibasic, probably a tertiary diamine, with the 
formula C 32 H 40 N 3 and bearing relationship to either berberine 
or cinchona (paricine). Some further properties which he at- 
tributes to the hymenodictyonine are these: 

DesaHption of the Alkaloid. — In the moist condition, as obtained by pre- 
cipitation with caustic alkali, it is a gelatinous mass of a cream colour and 
greedy of water, which it retains with extreme tenacity. By exposure it 
shortly acquires a decided yellow colour, which deepens with increase of 
temperature and passes into a light brown at 100°C. It has a persistently 
bitter taste, which is more quickly perceived when in solution than when 
in the solid state. It is readily soluble in alcohol, ether, chloroform, benzol, 
and light petroleum spirit. On evaporation of its ethereal solution at a 
slightly elevated temperature it separates out in the form of oily drops. 
If the heat be continued beyond that required for the complete evaporation 
of the ether, the oily drops coalesce and the whole assumes the character 
of a soft sticky resin. It commences to fuse at 66°C, and at 70°C. it 
will flow with ease sufficient to admit of its transference to another vessel. 
It neutralizes acids completely and the solutions are not fluorescent. It 
refuses to yield crystallizable salts with nitric, hydrochloric, acetic, sul- 
phuric, phosphoric and hydrobromic acids. Its solution in hydrochloric acid 
is precipitated by nitric acid, sodium nitrate and phosphate, potassium 
iodide, ferro and ferricyanide and bichromate, and mercuric chloride, in 
addition to the usual alkaloidal reagents. Potassium sulphocyanide added 
in excess to a neutral solution of the base in acetic acid gives reddish-yellow 
oily drops. A feebly acid solution gives with bromine a bright yellow 
precipitate and with solution of chlorinated lime a white precipitate un- 
affected by ammonia. A two per cent solution in 90 per cent alcohol is 
optically inactive. 

Naylor 4i did some work on Hymenodictyon excelsum prior to 
that already quoted. The difficulties encountered here were 
cleared up in the later work, in which radical changes were 
made in his methods of analysis and certain mistakes corrected. 

It is of interest to note that Naylor makes no mention of any 
fluorescence existing in any of his original solutions. In all the 
solutions from the material furnished in this investigation such 
a phenomenon was a marked factor in determining the relative 
extractive power of the solvent. 

Comparative tests on the products isolated from Philippine 
Hymenodictyon excelsum and on specimens of pure sesculin show 
them to be very similar in many of their reactions, and con- 
fusion as to the identity of the compound could easily arise. 

Some differences exist in the literature regarding the melting 
point of sesculin. 

"Naylor, W. A. W., ibid. (1883), 13, 817; (1884), 15, 195. 



xii. a. 4 Brill and Wells: Medicinal Plants, II 193 

Zwenger 15 claims that it melts at 160° C. with loss of water, 
and if heated over the melting point and then cooled, it crys- 
tallizes again free from its water of crystallization. 

Schitf 4U reports having had aesculin with 1.5 and with 2 
molecules of water of crystallization and that heating it above 
160- C. gives a compound which melts at 204° to 205° C. Heated 
to 230" C, the compound breaks down into sesculetin. Accord- 
ing to Schiff 47 aesculin loses part of its water of crystallization 
at 130° C. and the remaining 1.5 molecules at 160° to 165° C. 

The melting point is listed in Beilstein as 165° C. Melting 
points made by us on Kahlbaum's sesculin show sesculin to melt 
at 165° C, but immediately to solidify and then to melt at 200° 
to 205° C. with decomposition. In some cases where the heating 
was rapid no melting resulted at 165° C, only a slight softening 
was noticeable, and the only melting point observed in such 
cases was the higher one. 

The compound obtained from the Philippine Hymenodictyon 
excelsum by the various methods gives melting points varying 
from 195° to 205° C. No melting point or softening was ob- 
served at 165° C. A mixture of this compound with aesculin 
(Kahlbaum) melted at 169° C. and when heated to a higher 
temperature did not again solidify and then melt in the vicinity 
of 200° C, as does pure sesculin. It is much less soluble in 
water than is sesculin. It is not hydrolyzed to sesculetin by 
dilute hydrochloric or sulphuric acid as is sesculin. It is pre- 
cipitated by neutral lead acetate, 48 while sesculin is not. 

The compound was recovered from the lead precipitate by 
treatment with hydrogen sulphide. The melting point was sharp 
at 203° C. This is the original compound. No separation had 
taken place as we thought might be accomplished were the 
substance a mixture of sesculin and sesculetin, which seemed pos- 
sible from the reported work of Broughton 40 on the hymenodic- 
tyons. The compound is apparently a definite compound and 
not a mixture. It has properties which distinguish it from 
sesculin and from sesculetin, which is formed from sesculin by 
hydrolysis with dilute acids or emulsin. 

"Zwenger, C, Ann. d. Chem. (Liebig) (1854), 90, 65. 

48 Schiff, H., Ber. d. deutsch. chem. Gesell. (1881), 14, 303. 

47 Schiff, H., ibid. (180), 13, 1952. 

"' Our clarification of the extract from the plant by means of neutral 
lead acetate did not precipitate the compound, as only enough acetate was 
added to clarify the solution. 

43 Dymock, W., The Vegetable Materia Medica of Western India. 2d ed. 
(1885), 404. 



194 The Philippine Journal of Science mr 

The compound resembles sesculin by the fluorescence in solu- 
tion of carnelian yellow by transmitted light and of opalescent 
blue by reflected light. The intensity of the fluorescence is 
increased by the addition of ammonium hydroxide or other bases 
and is changed by the addition of acids to a purplish fluorescence 
of less intensity. Tests for nitrogen were negative. 

The compound is acid in solution and gives the blood red 
color produced by sesculin 50 when treated with nitric acid and 
ammonium hydroxide. A bromine product is formed when the 
compound is treated with bromine solution at room temperature. 
Crystallization was effected in acetic acid by warming on the 
steam bath. The product was readily soluble in acetic acid and 
difficultly soluble in alcohol. The dry substance reacts positively 
for bromine and negatively for nitrogen. The melting point 
is 235° C. with decomposition. The melting point of tribromses- 
culetin is 240° C. (not sharp, Beilstein), while monobromsesculin 
melts at 193° to 195° C. The compound may have been hy- 
drolyzed to sesculetin by the above treatment and tribromsescule- 
tin formed, or it may be the brom derivative of a different 
compound. A brom compound prepared by treatment of an 
acetic acid solution with bromine kept at the temperature of 
melting ice gave a compound with a melting point of 238° C. 

An acetyl derivative formed by treatment of the compound 
with acetic anhydride with gentle heating over a small flame 
had a melting point of 177° C. It is readily soluble in acetic acid 
and is thrown out by addition of water. When carefully crys- 
tallized, it is obtained in the shape of rhombic prisms and plates. 
The diacetyl derivative of sesculin melts at 130° C, and diacetyl 
aesculetin melts at 133° to 134° C. 

The compound isolated from the Philippine Hymenodictyon 
is neither sesculin nor sesculetin. We believe it to be identical 
with ^8-methyl-sesculetin found by Schmidt 51 in the roots of 
Gelsemium sempervirens. He describes /J-methyl-sesculetin as 
crystallizing in colorless, odorless, tasteless needles, which are 
soluble in ether, chloroform, and hot water. It gives a deep 
fluorescent solution with alkalies; a yellowish red solution with 
nitric acid, which changes to blood red on the addition of am- 
monium hydroxide and possesses a melting point of 202° to 
203° C. ; and a green color with ferric chloride. It reduces Fehl- 
ing solution when warmed. With a solution of gold chloride it 
gives a red, and with potassium permanganate a green, color. 

r,(1 van Rijn, op. cit., 288. 

"•'Schmidt Ernst, Arch. d. Pharm. (1898), 236, 324. 



xii, a, 4 Brill and Wells: Medicinal Plants, II 195 

By treatment with hydriodic acid according to the method of 
Zeisel, /3-methyl-sesculetin splits into methyl iodide and resculetin. 

The compound isolated from Hymenodictyon gives the above 
reactions, and we, therefore, announce the presence of /Mnethyl- 
aesculetin in the Philippine Hymenodictyon excelsum. 

SUMMARY 

The following-named plants were examined for physiologically 
active constituents with the results noted: Lophopetalum toxi- 
cum Loher contains a saponin which is poisonous in small 
quantities. Erythrophloeum densiflorum (Elm.) Merrill con- 
tains tannins, but no substances with any marked physio- 
logical properties. Quisqualis indica L. contains an oil in the 
seeds which has purgative properties ; a gum in the stems which 
is inactive physiologically, but gives some of the chemical tests 
of the alkaloids; and a considerable amount of potassium sul- 
phate. Tylophora brevipes (Turcz.) F.-Vill. contains an alka- 
loid identical in properties with tylophorine found by Hooper 
in Tylophora asthmatica growing in India. Toddalia asiatica 
(L.) Kurz contains the alkaloid berberine. Lunasia amara 
Blanco contains an alkaloid identical in properties with lunasine 
found by Boorsma in Lunasia costulata Miq. in Java. Rourea 
erecta (Blanco) Merrill is physiologically active toward the 
Carnivora, but inactive toward the Herbivora; the active prin- 
ciple could not be isolated, but further attempts will be made 
to isolate this when larger quantities of material are available. 
Hymenodictyon excelsum Wall. ( ?) contains /?-methyl-sesculetin. 
It differs from H. excelsum of India, since the latter contains 
sesculin, according to Broughton, while Naylor claims to have 
found an alkaloid, which has been named hymenodictyonine by 
him. 



ILLUSTRATIONS 

Plate I 

Fig. 1. Crystals of the hydrochloride of the alkaloid from Tylophora brev- 
ipes. 
2. Crystals of the free alkaloid from Toddah'a asiatica. 

Plate II 

Fig. 1. Crystals of the nitrate of the alkaloid of Toddalia asiatica from 
absolute alcohol. 
2. Same, crystallized from water. 

Plate III 

Fig. 1. Crystals of the free alkaloid from Lunasia amara by quick crystal- 
lization from alcohol. 
2. Same, crystallized from amorphous substance on long standing in 
vacuo. 

Plate IV 

Fig. 1. Crystals of the hydrochloride of the alkaloids from Lunasia amara. 
2. Crystals of the nitrate of the alkaloid from Lunasia amara. 

197 



Brill and Wells: Medicinal Plants, li.l 



[Phil. Journ. Sci., XII. A. No. 4. 




1 ffl 






i mm. 







Fig. 2. Crystals of the free alkaloid from Toddalia asiatica. 
PLATE I. 



Brill and Wills: Mi hkin at. Plants. II.] 



[Phil. Journ. Sci.. XII, A, No. 4. 






, 








Fig. 2. Same, crystallized from water. 
PLATE II. 



Brill and Wells: Medicinal Plants, II.] 



[Phil. Journ. Sci., XII, A, No. 4. 



,9 > ' ^ 



0J- 






yj 



Fig. 1. Crystals of the 






& 



d from Lunasia amara by quick < 




Brill ami Wills: MEDICINAL PLANTS, II.] 



[Phil. Joubn. Sci., XII, A, No. 4. 



4t % \M^i 




Fig. 1. Crystals of the hydrochloride of the alkaloids from Lunasia amara. 























Fig. 2. Crystals of the nitrate of the alkaloid from Lunasia amara. 
PLATE IV. 



THE ANTINEURITIC PROPERTIES OF THE INFUSORIAL EARTH 
EXTRACT OF THE HYDROLYZED EXTRACT OF RICE 
POLISHINGS » 

By Harvey C. Brill 

(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

The Bureau of Science prepares approximately forty liters of 
extract of rice polishing - for monthly distribution by the Liga 
Nacional para la Proteccion de la Primera Infancia for use in 
the treatment of infantile beriberi. 

In the past considerable difficulty has been experienced in pre- 
venting the fermentation of this solution after bottling. The 
practice of fractional sterilization at 70° C. prevents this fer- 
mentation, and the bottled extract remains sweet and unfer- 
mented indefinitely. But when the 50 cubic centimeter bottles 
of extract are distributed for use, fermentation often occurs due 
to the withdrawal of a portion of the contents, contamination 
with yeast cells, and the incubation at room temperature.- In 
an endeavor to prevent such fermentation, any one of several 
means might be employed: (1) A preservative might be added. 
Most, if not all, preservatives have a physiological action and 
the administration of them to people who are already in an un- 
healthy physical condition is to be avoided, if possible. (2) 
Smaller bottles of the extract might be distributed. There are 
several objections to this procedure. The cost of containers and 
the cost of handling would be greater; a smaller bottle would 
hold an amount insufficient to effect a cure or even improvement 
in some cases. The patient or the family of the patient would 
either need to make several trips to the dispensary or take more 
than one bottle of the extract. Either practice is to be avoided, 
if possible, on account of the difficulty of following the history 
of the case and the waste of extract that would result. (3) The 
extract might be issued in some form that would avoid the risk 
of fermentation. 

Infusorial earth has the power of extracting the salts of al- 

1 Received for publication April, 1917. 

2 The class afflicted with beriberi usually has no facilities for keeping its 
food and medicine cool. 

150276 3 199 



200 



The Philippine Journal of Science 



kaloids from acid solutions. 3 This property of infusorial earth 
caused Seidell ' to attempt the extraction of vitamine from brew- 
er's yeast.'' The above-named investigator obtained yeast from 
the breweries and subjected it to pressure to free it from the beer 
retained in it. The yeast was then autolyzed at a temperature 
of 37.5° C. and filtered. The brown filtrate was active, protecting 
a pigeon from the onset of polyneuritis for several months when 
administered in doses of 1 cubic centimeter on alternate days. 
This active liquid was treated with a prepared hydrous aluminium 
silicate (Lloyd's reagent) in the proportions and with the results 
given in Table I. 



Table I. — Extraction of yeast liquid with Lloyd's reagent. 



Quan- 
tity of 
yeast 
liquid. 


Quan- 
tity of 
Lloyd's 
reagent. 


Activity of filtrate. 


Activity 
of extract. 


cc. 

1.000 
1,000 
1,000 
1.000 


O. 

200 
20 
40 
50 




Active. 
Do. 






Do. 




Do. 







Twenty grams of Lloyd's reagent were insufficient to extract 
all the vitamine from a liter of yeast liquid ; 40 grams left a small 
proportion remaining in the filtrate. After extraction with 50 
grams of Lloyd's reagent, the filtrate possessed only slightly pro- 
tective powers from polyneuritis when administered to pigeons. 

EXPERIMENTAL PART 

Because of the difficulty experienced in keeping the extract of 
rice polishings sterile, it was decided to investigate the possi- 
bility of extracting the hydrolyzed extract of rice polishings 
with infusorial earth. This infusorial earth was obtained from 
Japan. Its absorptive powers are less than those of Lloyd's 
reagent. Under the conditions described by Lloyd for testing 
the efficiency of his reagent, Japanese infusorial earth reacted 
as follows: 



3 Lloyd, John U., Journ. Am. Pharm. Assoc. (1916), 5, 381. 

'Seidell, Atherton, U. S. Pub. Health Rep. (1916), 31, 364. 

6 Brewer's yeast contains the highest percentage of any of the known 
sources of vitamine. As it is a waste product in the breweries, it is also 
a cheap source of vitamine. 



Brill: Infusorial Earth Extract 



201 



Table II. — Absorptive power of Japanese infusorial earth for morphine 
bisulphate.* 



Quantity 

of in- 
fusorial 
earth 
used. 


Result with Mayer's reagent 
on the filtrate. 


a- 

0.60 
0.40 
0.35 
0.30 
0.25 


Not opalescent. 
Slightly opalescent. 

Do. 

Do. 
Strongly opalescent. 



° Five cubic centimeters of a 1 per cent solution of morphine sulphate diluted to 80 cubic 
centimeters. Lloyd reports his reagent to have an absorptive power of 8 grams of earth 
to 1 gram of morphine bisulphate. This is nearly the same proportion given by the Japanese 
earth. The latter has an absorptive power of 10 grams of earth to 1 gram of morphine 
bisulphate. 

The alcoholic extract c of rice polishings was hydrolyzed by 
heating with sulphuric acid in the manner described by Williams 
and Saleeby. 7 One 500 cubic centimeter portion of this hydro- 
lyzed extract was treated with 50 grams of the Japanese in- 
fusorial earth with shaking at intervals for a period of one hour. 
The mixture was allowed to subside, and the supernatant liquid 
was siphoned off. The infusorial earth was then placed on a 
suction funnel and washed with a 5 per cent sulphuric acid 
solution and finally dried. A second 500 cubic centimeters por- 
tion was treated with 25 grams of infusorial earth in a similar 
manner. The extracts, when dry, were placed in capsules for 
administration. The 50-gram sample, hereafter referred to 
as D, weighed 95 grams and filled 174 capsules, or each 
capsule represented the extract from 2.9 cubic centimeters of 
solution. The 25-gram sample, hereafter referred to as E, 
weighed 49 grams and was placed in 135 capsules, or each capsule 
represented the extract from 3.7 cubic centimeters of solution. 
The filtrate from D was bottled, subjected to fractional steriliza- 
tion, and marked B; the filtrate from E was bottled, subjected to 
fractional sterilization, and marked C. The original hydrolyzed 



' This extract is made by treating 25 kilograms of rice polishings with 
5 demijohns of 25 per cent alcohol and later pressing this. The extract is 
evaporated at a low temperature, the fat is separated, and alcohol is added 
to precipitate the albuminous and other matter, and again the extract is 
evaporated. The solution is concentrated, so that 1 cubic centimeter of 
extract is equivalent to 15 grams of rice polishings. 

'Williams, R. R.. and Saleeby, N. M., This Journal, Sec. B (1915), 10, 
106. 



202 



The Philippine Journal of Science 



extract of rice polishings is designated as A throughout this 
article. 

Two methods are applicable for the determination of activity 
in these extracts, namely, their protective property from the 
onset of polyneuritis in chickens fed on polished rice and their 
curative property when administered to chickens suffering from 
polyneuritis. Both methods were used for testing these extracts. 
Twelve chickens were placed on an exclusive diet of white rice 
and treated in the manner hereafter described. Two chickens 
had no treatment, but served as controls on the rice used. When 
polyneuritis became evident, they were treated with some one of 
the extracts, A, B, C, D, or E. Two chickens were given 3 cubic 
centimeters of A three times a week ; two received 3 cubic centi- 
meters of B three times a week ; two received 3. cubic centimeters 
of C three times a week ; two received one capsule of D three times 
a week ; while two received one capsule of E three times a week. 
The experiments extended over a total period of one hundred 
eighty-six days. Table III gives the history of the chickens fed 
on polished rice. These chickens had no treatment until polyneu- 
ritis had set in. They were then treated as described in the 
table until their condition improved, when the treatment was dis- 
continued. 

Table III.— History of control chickens. 



Days 

before 

onset of 

poly- 



of days 
on pol- 
ished 



3 cc. of A on alternate 

days. 
do 



Num- 
ber of 
treat- 
ments. 



1 capsule of D on alter- | 2 

nate days. 



Total 
loss in 
weight 



Results and remarks. 



Died at end of twenty-eight 
days. 

Improved; treatment discon- 
tinued. 

Improved ; treatment di scon- 
tinued. Living at end of 
experiment. 

Improved; gained weight af- 
ter treatment. Treat- 
ment discontinued. 

Improved; living at end of 
experiment. 



The weight of the chickens recorded in Table III decreased 
when polyneuritis set in. When treated, their weight increased 
at once. This increase amounted to as much as 20 per cent in 
some cases. 



Brill: Infusorial Earth Extract 



203 



In order that a check might be kept on the activity of the 
original extract of rice polishings, two chickens were treated 
with 3 cubic centimeters of A on alternate days. 

Table IV. — History of chickens treated with 3 cubic centimeters of A on 
alternate days (three times weekly). 



j Days Total I 
xi„ „* before number 
S onset of of days 1 
ch,ck - poly- onpol- 
neu- ished 
ritis. I rice. 



No extra treatment 

— .do 

do 



2 cc. of A daily . 

—do 

do 

-—do 



Num- 
ber of 
treat- 
ments. 



reg- 
ular. 
reg- 
ular, 
reg- 
ular. 
2 
3 
7 
10 



Total 
loss in 
weight. 



P.et. 

50 



Results and remarks. 



Died after refusing to eat 

for several days. 
Died without any symptoms 

of polyneuritis. 
Living at end of experiment 

in good health. 
Improved. 
Do. 
Do. 
Living at end of experiment, 

but in enfeebled condition. 



Three cubic centimeters of extract of rice polishings on alter- 
nate days were insufficient to prevent the onset of polyneuritis 
in chickens 1 and 4, but the extract prolonged the period of 
good health. Two cubic centimeters of A given daily for several 
days improved the condition of the chickens to such as extent 
that they could be again placed on the original treatment. 

To determine if 50 grams of infusorial earth would extract all 
the active substance from 500 cubic centimeters of hydrolyzed 
extract of rice polishings, the experiments recorded in Tables 
V and VII were planned. 

Table V. — History of chickens treated with 3 cubic centimeters of B on 
alternate days (three times weekly). 



No. of 
chick- 
en. 


Days 
before 
onset of 
poly- 
neu- 
ritis. 


Total 

number 
of days 
on pol- 
ished 


Treatment. 


Num- 
ber of 
treat- 
ments. 


Total 
loss in 
weight. 


Results and remarks. 


1 

2 
3 


101 
25 


104 

56 
186 


1 capsule of D on alter- 
nate days. 


2 

5 
none 


P.ct. 
46 

34 
34 


Was in serious condition 
before it was reported. 
Died at end of one hun- 
dred four days. 

Improved; living at end of 
experiment. \ 

Living at end of experiment. 







204 



The Philippine Journal of Science 



Of the three chickens treated one showed signs of polyneuritis 
at the end of one hundred one days, the second at the end of 
twenty-five days, while the third was still healthy at the end of 
one hundred eighty-six days. The vitamine has not been entirely 
extracted, as B still has protective properties. 

In the comparison made on morphine bisulphate, the Japanese 
infusorial earth compared favorably in absorptive powers with 
Lloyd's reagent, but Seidell found that 50 grams of Lloyd's re- 
agent extracted all the vitamine from 1 liter of autolyzed yeast 
liquid, while in the above experiment 50 grams of infusorial 
earth have not extracted all the vitamine from 500 cubic centi- 
meters of extract. The difference in the absorption must be due 
to the difference in the character of the two solutions. Hydro- 
lyzed extract of rice polishings is a thick, syruplike liquid, which 
undoubtedly collects on the infusorial earth and renders it in- 
capable of completely extracting the vitamine from the extract, 
while the yeast extract is no more viscous than water. 

When 500 cubic centimeters of extract of rice polishings are 
treated with 25 grams of Japanese infusorial earth, the filtrate, 
C, still has protective properties similar to those possessed by B. 



Table VI.— History of chickens treated with 3 cubic centimeters of C on 
alternate days (three times weekly). 



No. of 
chick- 
en. 


Days 

before 
onset of 
poly- 
neu- 
ritis. 

63 
63 


Total 
number 
of days 
on pol- 
ished 
rice. 

186 

68 

n. 


Treatment. 


Num- 
ber of 
treat- 
ments. 


Total 

loss in 
weight. 


Results and remarks. 


1 
2 


1 capsule of E on alter- 
nate days. 


1 

none 


P.ct. 
13 

41 

16 


Alive at end of experiment. 

Improved, but died four 

days later. 
In healthy condition at end 

of experiment. 







The infusorial earth extracts D and E possess antineuritic 
properties. The chickens, No. 1, of Table V, and Nos. 1 and 2, 
of Table VII, were benefited by the administration of D when 
they had contracted polyneuritis. Chickens 1 and 2, of Table 
VI, and 1, of Table VIII, were benefited by treatment with E. 

To test the activity of the infusorial earth extracts, the ex- 
periments described in Tables VII and VIII were performed. 



Brill: Infusorial Earth Extract 



205 



Table VII. — History of chickens treated with one capsule of D on alternate 
days (three times weekly). 



Days Total 
before number 
onset of of days 
poly- | onpol- i 
neu- ! ished 
ritis. rice. 



1 capsule of D daily 

do 

None 



Num- 
ber of 
treat- 
ments. 



Total 
loss in 
weight 



Results and remarks. 



Living- at end of experi- 
ments. 

Improved, but died fifteen 
days later. 

In healthy condition at end 
of experiments. 



Table VIII. — History of chickens treated with one capsule of E on alternate 
days (three times weekly). 



No. of 
chick- 
en. 


Days 
before 
onset of 
poly- 
neu- 
ritis. 


Total 
number 
of days 
on pol- 
ished 
rice. 


Treatment. 


Num- 
ber of 
treat- 
ments. 


Total 
loss in 
weight. 


Results and remarks. 


1 

2 
3 

4 


105 
81 


122 

83 

81 

100 


1 capsule of E daily 


none 
none 

none 


P.ct. 
23 

46 


Improved, but died ten days 

later. 
Alive at end of experiments. 
Died without any extra 

treatment. 
Alive at end of experiments. 




do 





DISCUSSION 

The results obtained show that the curative principle is ex- 
tracted from the hydrolyzed extract of rice polishings by means 
of infusorial earth, although not by as small a proportion of the 
earth as was found by Seidell when he used Lloyd's reagent with 
autolyzed yeast liquor. This partial extraction is undoubtedly 
due to the character of the extract. 

Extract A, given on alternate days in doses of 3 cubic centi- 
meters, did not protect the chickens from polyneuritis for the 
entire period of one hundred eighty-six days. Chicken 1 con- 
tracted polyneuritis on the eighty-seventh day, chicken 2 died 
on the seventy -third, while 4 became sick on the fifteenth day of 
its confinement or the one hundred thirty-first day of the experi- 
ment. An administration of 3 cubic centimeters of A on alter- 
nate days should be sufficient to protect these chickens. It is 



206 The Philippine Journal of Science 

possible that this extract no longer possesses its full curative 
properties when it ages. Treatment of chicken 4 was begun- 
on the one hundred sixteenth day of the experiment ; consequently 
the extract used had been made for some time. Williams 8 has 
recently shown that hydroxy pyridines have antineuritic proper- 
ties, but these curative effects are lost after a solution stands for 
some time or more quickly when warmed. Evidence that the 
antineuritic compounds in autolyzed yeast undergo a change 
probably due to dynamic isomerism has been presented by Wil- 
liams and Seidell. 9 A preparation from cardiac muscle loses its 
ameliorative power in a few days. 10 A similar change may be 
the cause of the apparent lessened activity with age of the 
hydrolyzed extract used in these experiments. 

The method of extraction with infusorial earth appears in- 
applicable, due to the inability of small quantities of infusorial 
earth to absorb the total vitamine content of the hydrolyzed 
extract of rice polishings. 

SUMMARY 

A study of the antineuritic properties of infusorial earth ex- 
tracts of the hydrolyzed extract of rice polishings is reported. 

Only a part of the vitamine content of the extract was ex- 
tracted by the proportions of infusorial earth used. 

There appears to be a loss of antineuritic power in the extract 
as it ages. 

8 Williams, R. R., Journ. Biol. Chem. (1916), 25, 437. 

8 Williams, R. R., and Seidell, Atherton, ibid. (1916), 26, 431. 

10 Cooper, Evelyn A., Biochem. Journ. (1914), 8, 347. 



THE USE OF CHAULMOOGRA OIL AS A SPECIFIC FOR LEPROSY l 

By Harvey C. Brill and Robert R. Williams 
(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

Regarding the cure of leprosy by means of chaulmoogra oil, 
considerable differences of opinion exist. Thompson, 2 of New 
South Wales, was unable to effect any cures by its use. The 
Philippine Health Service released twenty-three cases from Cu- 
lion leper colony about a year ago that had been treated with 
chaulmoogra oil. No positive claims that it brought about these 
cures are made by the more judicial members of the Health 
Service staff, nor can these patients be considered permanently 
cured until they have been under observation for a longer period 
of time. However, chaulmoogra oil appears to be the most pro- 
mising remedy for the treatment of leprosy at present known. 

The administration of chaulmoogra oil is accompanied by con- 
siderable difficulties. When administered by mouth, nausea and 
digestive disturbances follow. Given subcutaneously, it is ab- 
sorbed slowly, since it is a heavy oil; consequently its adminis- 
tration in this manner is painful, due to the pressure on the nerves 
and the sores that result from the slow absorption. Then not all 
cases react to the treatment. The nonuniformity of the results 
that follow its administration may be due to a difference in the 
quality of the oil or to a variation in the disease itself in dif- 
ferent localities. 

Crude chaulmoogra oil is usually recommended for use. s 

The recommendations of those engaged in its administration 
are strongly in favor of the crude oil, but these opinions may 
partly be a matter of prejudice. These practicioners undoubt- 
edly have been previously disposed in its favor by their reading, 
and as observations of leprosy treatment are not very general 
nor very decisive, and as recovery in any case is slow, decisions 
as to its efficiency are difficult to form. We must conclude that 
cases of positive cures by means of chaulmoogra oil are not de- 
finitely proved. 

1 Received for publication January 29, 1917. 

2 Report of the Board of Health on Leprosy in New South Wales (1908). 
'United States Dispensatory. 18th ed. (1899), 1678. 

207 



208 



The Philippine Journal of Science 



EFFECTIVENESS OF CHAULMOOGRA OIL 

Again, standards for determining whether an oil is crude or 
refined are rather arbitrary, and when applied to an oil such as 
chaulmoogra, where the previous history of the oil is not known, 
the difficulty of deciding is still more difficult. An expressed oil 
would carry with it other substances. For example, when the 
oil from bitter almonds is expressed, a small amount of the amyg- 
dalin of the nuts is carried with the oil. Power and Lees * have 
shown that the chaulmoogra seeds and the hydnocarpus seeds all 
contain small quantities of a cyanogenetic glucoside that they 
found to be identical with gynocardin, the glucoside present in 
Gynocardia odorata. 

When the oil of these nuts is expressed, minute amounts of 
the glucoside doubtless accompany the oil ; therefore the effective- 
ness of the crude oil may possibly be due to the presence of the 
cyanogenetic glucoside. The small amount of the glucoside 
present would account for the slowness of the action of the oil. 
The small amount of glucoside that may possibly be present in 
the oils is shown by the nitrogen content of some chaulmoogra 
oils in Table I. These data illustrate the maximum possible 
amount that could be present. 

Table I. — Nitrogen content of various chaulmoogra oils. 



No. of sample. » 


Nitrogen. 


Calculated as 
grynocardin 
(CisHisOsN 
+1.5H»0). 




Per cent. 
0. 01363 
0.00438 
0.01380 
0.01415 
0.01257 
0.00670 
0.01503 
0.02209 


Per cent. 
0.350 
0.113 
0.355 
0.364 
0.323 
0.172 
0.386 
0.568 










6 









For description of samples see Table III. 



CONSTITUTION OF THE OIL 



The acids of the oils of this family of seeds, Taratogenos kwr- 
zii, 5 Hydnocarpus ivighttiana, H. anthelmintica, 6 H. venenata, 



1 Journ. Chem Soc. London (1905), 87, 349. 

'Power and Gornall, Journ. Chem. Soc. London (1904), 85, 838. 

'Power and Barrowcliff, ibid. (1905), 87, 884. 

'Brill, This Journal, Sec. A, (1916), 11, 75. 



xii, a. 4 Brill and Williams: A Specific for Leprosy 209 

H. alcalae, and Pangium edule, 9 are peculiar in that they are 
monobasic unsaturated acids with a ring structure. 1 ' Chaulmoo- 
gric acid is isomeric with linoleic acid (C 8 H 82 2 ), but differs 
from the latter in that it combines with but two bromine or iodine 
atoms and contains but one ethylenic linking and a closed chain. 
It is very stable toward alkalies. Heating to 300° C. with alkalies 
does not decompose it. It is optically active, rotating the plane 
of polarized light to the right. These properties make it dis- 
tinct from all other known acids, and one would expect its phys- 
iological properties to be different from those of other acids. 

Hydnocarpic, the other peculiar acid found in the oil of these 
seeds, is a lower member of this same series. 

Chattopadhyay 10 has taken exception to the conclusions of 
Power and his coworkers regarding the presence of acids of the 
nature of chaulmoogric and hydnocarpic in chaulmoogra oil, but 
his results are not complete enough to substantiate his contention. 
The results by one of us (Brill) on H. venenata 11 are in agree- 
ment with those of Power et al., and unless Chattopadhyay can 
produce results which really contradict those of Power, the results 
of the latter will continue to be accepted. That the effectiveness 
of chaulmoogra oil may be due to the presence of free acids has 
received further confirmation recently by the experiments of 
Leonard Rogers, 12 who administered the free acids themselves 
and these in the form of their sodium salts. However, the com- 
pound known as antileprol, which is the mixture of the ethyl 
esters of chaulmoogric and hydnocarpic acids, was without effect 
when used by the Philippine health authorities. We should ex- 
pect the ethyl esters to have an effect as powerful as that of the 
free acids, since they would be readily hydrolyzed in the system. 
Indeed it is hardly to be expected that the free acids should be so 
radically different in activity from the glyceryl esters, the neutral 
oil itself, since the oil when absorbed would be undoubtedly 
hydrolyzed to some extent in the system. The use of the soluble 
alkali salts would appear to be more promising on account of 
their solubility and more ready absorption, and as Rogers obtains 
promising reactions when a solution of these salts is administered 
intravenously, they deserve to be tested thoroughly by the va- 
rious scientists interested in the treatment of leprosy. 

The work described in this article was undertaken in the hope 

s Brill, ibid. (1917), 12, No. 1. 

' Barrowcliff and Power, Journ. Chem. Soc. London (1907), 91, 557. 

10 Chattopadhyay, P. C, Am. Journ. Pharm. (1915), 87, 473. 

u Loc. cit. 

12 Lancet (1916), 190, 288; Indian Med. Gaz. (1916), 51, 195 and 437. 



210 



The Philippine Journal of Science 



that some light might be thrown on the constituent in chaulmoo- 
gra oil that is specific for leprosy; that it might be isolated and 
that its isolation would simplify the administration of the oil ; and 
that the more concentrated form of the physiologically active 
substance would bring about speedier improvements in the con- 
dition of the patients. While the results have not been entirely 
successful in this endeavor, yet in the hope that they may give 
an inspiration to other investigators, and because of the infor- 
mation they give concerning chaulmoogra oil itself, they are 
presented here. 

The administration of the preparations was kindly performed 
by the Philippine health authorities at the Culion leper colony. 

Circumstances were such that in some cases the treatment was 
not continued over a sufficiently prolonged period, nor were the 
administrations frequent enough. Denney 13 has described the 
difficulties of persuading lepers to continue a treatment for any 
length of time. For this reason some of the fractions hereafter 
described might show more satisfactory results under more 
favorable circumstances. The treatment at Culion is entirely 
voluntary. The leper voluntarily presents himself for treat- 
ment, and the treatment is continued only as long as the patient 
is in the humor for taking it. No coercion can be applied ; con- 
sequently the health officer is handicapped in his desire for a 
systematic study of the treatment of leprosy. 

Table II. — Constants of various chaulmoogra oils examined by the Bureau 
of Science. 



Sample No. 


Specific 
gravity 
at 30° C. 


Specific 
rotation 
in chloro- 
form for 

sodium 
light. 


Acid va- 
lue cc. 
0.1 N 
base. 


Saponifi- 
cation 
value. 


Index of 
refrac- 
tion. 


Iodine 
value. 




0. 9535 
0.9464 
0.9492 
0.9492 
0.9429 
0.9487 
0.9484 
0. 9467 
0. 9470 
0.9454 


+62.23 
+ 52.12 
+45. 69 

+57.45 
+48.95 
4 58.20 
+48. 73 
+46.26 
+ 51.60 
+47.30 


2.69 
7.19 

21.48 
5.63 
1.55 
6.36 
2.80 
5.56 


196.6 
189.1 
191.6 
196.8 
196.3 
196.2 
210.5 
210.5 
215.0 


1.4755 
1.4768 
1. 4766 
1. 4730 
1. 4720 
1.4732 
1. 4755 
1. 4750 
1. 4774 
1. 4762 


99.3 
101.1 
102.6 

79.6 
100.1 
102.1 
110.4 
104.6 
107.5 

99.5 


















10. __ 




0. 9535 
0. 9429 


+58.20 
+45.69 


21.48 
1.55 


210. 5 
189.1 


1. 4774 
1. 4720 


110.4 
97.6 


Minimum -— 




0.9471 


+50.81 


6.79 


200.4 


1.4751 


102.4 





Denney, Oswald E., This Journal, Sec. B (1915), 10, 357. 



xii. a. 4 Brill and Williams: A Specific for Leprosy 211 

For comparative purposes the constants of chaulmoogra oil 
used in this experimental work and of other samples of chaul- 
moogra oil that have been examined in the Bureau of Science are 
inserted. 

DESCRIPTION OF SAMPLES OF CHAULMOOGRA OIL 

Sample 1. — This oil was obtained from the Philippine Health 
Service from the stock used by them in the treatment of the 
lepers at Culion. It was a greenish yellow limpid oil (30°C.) 
with a characteristic odor. 

Sample 2. — This oil was extracted from seeds by means of 
petroleum ether. The oil was liquid (30° C.) ; it was somewhat 
lighter in color than sample 1, but had the same odor. The 
seeds were obtained from H. G. Carter, economic botanist to 
the Botanical Survey of India, Calcutta, India, and were labeled 
"chaulmoogra seed — Taraktogenos kurzii." 

Sa.viple 3. — This is an oil purchased from the German Dis- 
pensary, Manila. The oil was of the same color and odor as 
sample 1, but differed from the preceding in that it had a small 
quantity of solid particles deposited, about two parts per 
hundred. 

Sample U. — This is a sample of oil purchased from the Botica 
de Santa Cruz, Manila, and marked "Aceite de Chaulmugra." 
The sample was practically identical with sample 1 in appearance 
and other properties. 

Sample 5. — This sample was purchased from the Botica de 
Santa Cruz. It bore the following on its label: 

"Chaulmoogra oil." The fixed oil expressed from the seeds of Gynocardia 
odorata R. Brown natural order Bixaceae. Should the oil become solidified 
from cold, it may be readily liquified by placing the container in warm water. 

Dose 10 to 20 minims (0.6 to 1.3 cc). Guaranteed under the Food and 
Drugs Act, June 30, 1906, guaranty No. 6. 

This oil was so nearly solidified at the ordinary temperature 
(30° C), that it would not readily pour. 

Sample 6. — This oil was submitted by the Philippine Health 
Service for test to determine its purity. The sample was liquid 
and possessed the odor and color of sample 1. 

Sample 7. — This oil was expressed from seeds purchased 
through the Department of Agriculture, Assam, Seed Depot 
Gauhati. They were marked "gynocardia odorata," but were 
sold to the Bureau of Science in response to a request for chaul- 
moogra seeds. The oil was a heavy, limpid oil, darker in color 
than sample 1, but with the same characteristic odor. 

Sample 8. — This oil was submitted by the Philippine Health 



212 The Philippine Journal of Science 1917 

Service for testing as to purity. It was marked: "Finest oil 
of Chaulmoogra, Stanistreet Brand." It is very similar in prop- 
erties to sample 7. 

Sample 9. — This oil is identical in all respects with sample 7, 
being obtained by expression from seeds from Assam, marked 
"Gynocardia odorata." 

Sample 10. — This oil was submitted by the Philippine Bureau 
of Supply for testing as to purity. It was marked: "Crude 
Chaulmoogra Oil. A. S. Watson, Hongkong China." The oil 
was very similar to sample 1. 

A survey of the constants of the oils described above shows 
that they are different from the constants given by Power and 
BarrowclifF 14 for the oil from Gynocardia odorata and that the 
name Gynocardia odorata used for some of the oils examined 
by us is incorrect, but that they are in substantial agreement 
with their results for the chaulmoogra and the hydnocarpus oils. 

CHEMICAL EXAMINATION OF CHAULMOOGRA OIL 

A more thorough chemical examination of chaulmoogra oil 
was made for the purpose of substantiating the results of either 
Power or those who disagree with him. For this examination 
samples 2 and 5 were chosen. They were saponified with alco- 
holic potash, and the free acids were obtained by acidification 
and extraction with ether. Table III gives the properties of 
the free acids. 



Table III. — Properties of the free acids of chavlmoogra oil. 




Sample No.— 


2 1 6 


Specific rotation in chloroform for sodium light at 30°C _ _- 


+54.00 ! +47.40 
101.3° | 99.7° 
"35.31 a 34. 95 
38-39 ! 3S-40 









These acids were treated in the manner adopted by Power 
for the separation of chaulmoogric and hydnocarpic acids. 

Not much difficulty was experienced in obtaining a body from 
No. 5 that melted at 67° to 68 °C. when recrystallized from al- 
cohol, but considerable difficulty was experienced in purifying 
the fraction from No. 2 sufficiently to raise its melting point to 

" Journ. Chem. Soc. London (1905), 87, 887. 



Brill and Williams: A Specific for Leprosy 



213 



67° to 68 c C. The properties of these acids are tabulated in 
Table IV. When the residue of acids after the separation by 
means of crystallization from alcohol was treated with barium 
acetate, four fractions were obtained. These had the melting 
points noted in Table IV. 



Table IV. — Melting point of acids obtained 
acetate. 



treatment with barium 





Sample No.— 




2 


5 




°C. 
36-37 
40-41 
39-40 


°C. 

47-48 
38-39 
39-40 








37-38 1 34-35 









Note. — Fractions 2 and 3 were in both cases united, and the combined portions were 
reerystallized from alcohol, discarding the portion first crystallizing out. They were then 
again converted into the barium salts, the acids set free, and again crystallized from alcohol. 
The melting points correspond to the melting point of hydnocarpic acid, 59° to 60° C. 

Table V. — Saponification values of acids isolated from chaulmoogra oil. 



Sample. 


O.liVbase 

1 required 

Quantity.' f or neu . 

| traliza- 

tion. 


Saponi- 
fication 
value. » 




g. cc. 


200.5 
202. 4 
222.7 
218.2 




1. 0587 
1. 9205 
2.8560 


38.13 
76.21 
110. 82 









"Theoretical saponification value for chaulmoogric acid, C n H 31 COOH = 200.5 ; for hydno- 
carpic acid, C, 6 Hi, COOH = 222.7. 

Table VI.— Specific rotation in chloroform at 30° C/D of acids isolated 
from chaulmoogra oil. 



Sample. 


Quantity 
in 25 cc. 
chloro- 
form. 


Rotation 
in cm. 

tube. 


Specific 
rotation. * 


Chaulmoogric acid: 


a- 

1. 8120 

1.4416 

2. 6740 
1.5545 


Degrees. 

+2.14 
+ 1.68 

+7.24 


Degrees. 

+59.05 
+58. 10 

+67.70 


5...- __ . 


Hydnocarpic acid: 

2. 


5 









Found by Power for chaulmoogric acid, 58.6° ; for hydnocarpic acid, 68.1°. 



214 The Philippine Journal of Science 1917 

Table VII. — Iodine value of acids isolated from chaulmoogra oil. 



Sample. 


Quantity. 


0.1 A' 
iodine 
solution 
used. 


Iodine 
value 
(Han- 

U3).» 


Chaulmoogric acid: 


a- 

0.1806 
0. 1558 

0. 2250 
0. 1625 


12.73 
11.13 

17.71 
12.73 


89.5 
90.7 

100.2 
99.9 




Hydnocarpic acid: 


::::::;::::;;;:;:: :;::;: :::;;:::::::: .:::::::: 


i 



1 Theoretical iodine value for the addition of two atoms of iodine to chaulmoogric acid, 
Ci 7 H 3l COOH = 90.6 ; for the addition of two atoms of iodine to hydnocarpic acid. C15H07COOH 
= 100.6 

Preparation and description of samples. — A sample of chaul- 
moogra oil was washed with three separate portions of alcohol. 
The alcohol took on the color of the chaulmoogra oil, and a par- 
tial solution of the two took place. The separation was effected 
by adding water to the mixture, when the alcohol and oil formed 
separate layers. The alcohol-washed oil had the properties 
noted in Table VIII. 

Table VIII. — Properties of chaulmoogra oil after washing with alcohol 

Specific gravity 0.9530 

Specific rotation in chloroform for sodium light +52.11 

Acid value cc. 0.1 N base 2.46 

Saponification value 196.3 

Index of refraction 1.4735 

The alcohol extract after extraction with petroleum ether was 
distilled under lowered pressure until all the water and alcohol 
were removed and only a small residue remained; it was then 
dried over sulphuric acid in a vacuum desiccator. The residue 
is a golden yellow, viscous, almost odorless, noncrystallizable 
mass. It constitutes about 0.06 per cent of the original oil. It 
was insoluble in olive oil, but was suspended in 50 cubic centi- 
meters of this oil, marked No. 1, and submitted to the authorities 
at Culion with the following directions: 

Directions: 1 cubic centimeter = 80 cubic centimeters chaulmoogra oil; initial 
dose 0.10 cubic centimeter increasing to 0.50 cubic centimeter; shake 
before using. 

The sample used in this manner at no time caused a reaction, 
that is, it appeared to be physiologically inactive when given 
in the above doses. The petroleum ether extract of the alcohol 
extract was placed on the steam bath for the removal of the pe- 
troleum ether and finally dried in a vacuum desiccator. This oil, 



Brill and Williams: A Specific for Leprosy 



215 



marked No. 2, is darker than No. 3 (described later) ; it is solid 
at ordinary temperature (30° C). It constituted about 26.2 per 
cent of the original oil. 

Directions: Dose same as chaulmoogra. 

The alcohol-washed oil was then placed in a cold room (tem- 
perature, 17" to 19° C.) for four days, where it hardened to a 
viscous mass. When placed in cheesecloth and hung, only a 
very small portion separated. It was then placed in a hand 
press and subjected to pressure. The fractions resulting were 
again subjected to fractionation by means of the press, and the 
hard fractions were mixed, and the soft samples that resulted 
were mixed. The soft oil fraction, known as No. 4, was the 
darkest of the three samples. It is soft at 17° C. and constitutes 
about 20.7 per cent of the original oil. 

The hard oil fraction, submitted to Culion as No. 3, is darker 
than No. 2, but not so dark as No. 4. It is hard at 17° C. and 
constitutes about 53 per cent of the original oil. The directions 
for Nos. 3 and 4 were identical with the directions for No. 2. 

The chemical constants of these three fractions are tabulated 
in Table IX. 

Table IX.— Constants of fractions Nos. 1, 2, and 3. 



Sample No. 


! Specific 
c, !a _ ' rotation 

at du" u. sodjum 
light. 


Acid va- 

luein cc. 

0.1 N 

base. 


Saponifi- 
cation 
value. 


Index of 
refrac- 
tion. 


Iodine 
value. 


Original 
oil. 


1 
Remarks. 


2. 


0.9440. +52.35 
0.6522 1 +55.03 
0.9537 -54.84 


17.08 


190.2 
194.4 
191.3 


1.4699 




P. cent. 
26.2 


Solid at 30° C. 
Solid at 17° C. 
Liquid at 17° C. 






4 




1.4764 


103.4 


20.7 



None of the fractions showed any greater activity than the orig- 
inal chaulmoogra oil, that is, no more marked reactions were ob- 
served than are caused by the administration of chaulmoogra oil. 

Sample 2, Table II, which was obtained by petroleum ether 
extraction of seeds received from Madras from H. G. Carter, 
economic botanist to the Botanical Survey of India, and listed 
as seeds of chaulmoogra, was forwarded as No. 5. Its proper- 
ties place it with chaulmoogra oil. Directions: Dose same as 
chaulmoogra oil. The sample gave no reaction, differing in in- 
tensity from that of ordinary chaulmoogra oil when it was ad- 
ministered. It seemed no less active than many commercial 
samples used by the Public Health Service. 



216 The Philippine Journal of Science 

Table X. — ■Data on seeds of chaulmoogra. 





Grams. 


Per cent. 


Whole nu 


10. 366 
6,372 
3.994 
5,959 
3.160 




Kernels 


61.47 
38.63 
57.48 










30.48 
53.03 








370 




3.57 
6.20 








350 




3.37 
5.87 











The ground petroleum ether extracted nuts were then extracted 
with hot absolute alcohol, and the alcohol extract was evaporated 
to small bulk. A large quantity of sugar (glucose) separated 
from the mother liquor. After the separation of the free glucose 
the extract was evaporated to dryness, and the residue was dis- 
solved in water and extracted with chloroform. The chloroform 
was evaporated, and the residue was dissolved in water and 
evaporated very slowly over sulphuric acid. No crystalline prod- 
uct was obtained. The sample had a somewhat bitter taste. 
It was dissolved in a mixture of alcohol and olive oil and for- 
warded to Culion as No. 6. 

Directions : Contents, 100 cubic centimeters; initial dose 1 cubic centimeter, 
to be increased later if advisable; shake before using. 

In one case this sample gave a decided reaction, but no repeti- 
tions of this reaction were obtained. 

One half of the material after having been extracted with 
chloroform was dissolved in 200 cubic centimeters of 15 per cent 
alcohol and forwarded as No. 7. 

Directions: Initial dose 1 cubic centimeter to be increased later if advisable; 
shake before using. 

No noticeable reaction was obtained with this fraction. 

An attempt was made to isolate a crystalline product from 
the remaining half of the extract. A small amount of sugar 
only was obtained. Apparently the glucoside, originally pres- 
ent, has been hydrolyzed by the enzymes and moisture. 

One half of the water extract obtained by treatment of the 
ground seeds after the alcohol treatment was dissolved in 500 
cubic centimeters of 15 per cent alcohol and submitted as No. 8. 



xii. a.i Brill and William*: A Specific for Leprosy 217 

The remainder was examined for crystalline bodies without any 
success. 

Directions: Initial dose 5 cubic centimeters, to be increased later if ad- 
visable; shake before using. 

No reaction was observed when No. 8 was administered. 

The ethyl esters of the acids present in chaulmoogra oil were 
prepared by esterification of a mixture of alcohol and chaul- 
moogra oil, using hydrochloric acid gas as the dehydrating agent. 
The excess acid was neutralized by adding dry sodium carbonate. 
This sample was submitted as No. 9. No observed reactions 
attended its administration. 

Because of the reputed superiority of the crude oil over the 
refined oil, the two remaining samples were prepared. To de- 
termine if the greater effectiveness of the crude oil might be due 
to the presence of free acids, No. 10 was prepared. Two hun- 
dred grams of chaulmoogra oil were saponified, and the free 
acids were dissolved in a portion of the original oil. This pre- 
paration had an acid value of 11.47 cubic centimeters 0.1 N base. 
Directions: Dose same as chaulmoogra oil. 

The reactions observed were reported as no more marked than 
the reactions caused by chaulmoogra oil itself. 

EXPERIMENTS WITH AMYGDALIN 

The other speculation, regarding the cause of the greater 
effectiveness of the crude oil, that it might be due to the pre- 
sence of small amounts of the cyanogenetic glucoside, induced 
us to investigate the effect of amygdalin on leprosy. Amyg- 
dalin ls is hydrolyzed by emulsin, crab juice, and acids, giving 
glucose, benzaldehyde, and hydrocyanic acid. Wchler and Fre- 
richs Ui state that amygdalin in small quantities is not poisonous 
to dogs, but when given in quantities of from 4 to 5 grams it 
causes sickness, and when accompanied by emulsin it is extremely 
poisonous, since it is hydrolyzed by the enzyme giving free hydro- 
cyanic acid. 

To determine the rapidity with which it is eliminated from 
the blood stream, 0.25 gram was dissolved in 1 cubic centimeter of 
water, and this solution was injected into the vein of the right 
ear of a rabbit. Four rabbits were so treated. At the end of 
two, ten, and twenty-four hours 0.5 cubic centimeter quantities 
of blood were drawn from the left ears of the rabbits, and this 

15 Abderhalden, Emil, Biochemisches Handlexikon. Julius Springer, Ber- 
lin (1911), 2, 707. 

"Ayinal. d. Chem. & Pharm. (1848), 65, 337. 



218 



The Philippine Journal of Science 



blood was tested for the presence of amygdalin by hydrolysis 
with hydrochloric acid and was tested for hydrocyanic acid 
by distilling into alkali and using ferrous sulphate paper, with 
the results noted in Table XI. 



Table XI. — Elimination of amygdalin from blood stream of rabbits. 



Rabbit. 


Results. 


2 hours. 


10 hours. 


24 hours. 


1 


+ + 
+ + 
+ + 
+ + 


(») 

w 

(») 
(") 


" 


2 


8 


4 





++, fair test for hydrocyanic acid ; 
■ Questionable. 



negative for hydrocyanic acid. 



The results indicate that the amygdalin no longer exists in 
the blood stream after ten hours. 

Tests were made with blood and blood serum from lepers and 
with blood from apparently healthy people by incubating the 
samples with a standard amygdalin solution to determine if any 
difference exists in their hydrolytic powers on amygdalin. 

Table XII. — Quantity of hydrocyanic acid liberated by the action of blood 
on amygdalin. 



Kind of blood. 



Quantity 
of amyg- 
dalin solu- 
tion (1 
gram in 
150 cc. of 
water). 



Quantity of 

0. 1 N silver 

nitrate 

solution 

used. 



Hydro- 
cyanic acid 

liberated 
in per cent 
of original. 



Hydrocy- 
anic acid 
liberated. 



Leper 

Blood serum from leper 

Healthy 

Do. 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

Do 

None 



0.47 
9.49 

0. 34 



3. 68 
5.50 

0.60 

1. 05 
1.06 

0.77 
1.80 

1.33 
0. 94 
1.05 
1. 10 



0.000540 
0. 000059 
0. 000059 
0. 000103 
0.000103 
0.000097 
0. 000076 
0. 000128 
0. 000132 
0.000092 
0. 000103 
0. 000108 



While the absolute results may not be of extreme accuracy, 
they are comparable and indicate that leper blood may have 
greater hydrolytic power on amygdalin than does normal blood. 



xii, a. 4 Brill and Williams: A Specific for Leprosy 219 

It is to be regretted that we were unable to procure more leper 
blood in order that this study might have been extended. 

To determine the action of amygdalin on lepers, tablets weigh- 
ing approximately 0.05 gram were made and submitted as No. 
11. 

Directions: One tablet every other day gradually increasing the size and 
frequency of the dose. 

The Culion authorities reported no marked reactions with 
amygdalin, but the trial was not over a long period. It might 
give more favorable results under improved conditions. 

HYDNOCARPUS ANTHELMINTICA 

For the purpose of further confirming the results of Power 
and others on their work on chaulmoogra and hydnocarpus 
oils, the constants of the oil from the seeds of Hydnocarpus 
anthelmintica obtained from Madras, through the kindness of 
H. G. Carter, and examined in the Bureau of Science, are added. 

Table XIII. — Constants of oils from Hydnocarpus anthelmintica seeds. 





Results. 




0.9487 

+49.50 

0.6 

206.2 

1.4725 

90.8 


» 0. 9520 

51.0 

8.1 

208.0 












82.5 





a At 25° C. Results of Power and Barrowcliff, loc. cit. 

The results are close enough to prove the identity of the 
seeds furnishing the two oils. 

DISCUSSION 

As has been previously stated, the medicinal administration of 
chaulmoogra oil should be accompanied by the chemical control 
of the samples used. It seems plausible to think that the slow- 
ness of the changes caused by the use of chaulmoogra oil in the 
treatment of leprosy may be due to the small quantity of the 
active constituent present in the oil. If this constituent could 
be isolated and administered in concentrated form, more rapid 
cures should result, and the treament would undoubtedly be less 
painful. 

The results of Rogers 17 with the free acids and the sodium 

" Loc. cit. 



220 The Philippine Journal of Science 

salts make the use of these appear as promising remedies. But 
Rogers cautions against the assumption of a too optimistic at- 
titude, as he regards the problem as still unsolved. On the 
other hand, it seems likely that antileprol and the neutral oil 
should be more effective than they have been found to be, if 
cures result from the use of the free acids and of the sodium 
salts. Consequently the inactivity of antileprol and of many of 
the commercial chaulmoogra oils should make practitioners cau- 
tious about accepting a remedy as specific for leprosy until it is 
proved to be such. 

SUMMARY 

The constants of ten samples of chaulmoogra oil are given. 
A chemical examination of chaulmoogra oil is included. The use 
of various fractions of chaulmoogra in the treatment of leprosy is 
discussed, and the desirability of chemical investigation accom- 
panying the medicinal administration is pointed out. 






PUBLICATIONS rOK SALE BY THE BUREAU OP SCIENCE, 
MANILA, PHinPPINE ISLANDS— Continued 















A FLOKA OF MANILA 














. 419 


Paper, 


490 


page* 


, $2.50. 










ra of the cul- 
ines. Descrip- 
species, 590 
native names, 



FEiLiPPnrE difxeugcaxp toi 



150 pages, 1 ma a. 
L2 diaorams, $1.00, 



ZOOLOGY— Contlnned 

A 5CANTT AX OF PHILIPPINE BIEDS 



$4, postpaid. 

i Manual of Philippine Birds con 
lompact form descriptions of all 
fri species of Philippine birds. The 
and diagnoses of orders, families, 
ra help the novice in identification. 



A CHECK-LIST OF PHILIPPINE 



arocarp Forest3 the 
ve discus- 
development of dip- 
he other elements of 



rALATAN WOODS 



stpa 



This list will be found a convenient guide 
to the synonymy of Philippine ichthyology. 
The nomenclature is thoroughly revised, and 
the distribution of each specie6 within the 
Philippine Islands is given. 



MKALS OP 
LANDS. EXCLi 
[£ CETACEA 



■ No. 436. Paper. 195 pages, $1.C0. 
postpaid. 

his catalogue includes the rcmes of all 
is of Coleoptera that have been recorded 
locality in the Philippine 
Is. References to original descriptions 
other important notes are given. The 
imic appendix includes oommertt on 
species of beetles which are known to 
jurious or beneficial to rrian. 



VTES CURRENCY 



PHILIPPINE J0L r : 
or to any of the ; 



le BUSINESS MANAGER, 
SCIENCE,' MANILA, P. I., 

:-' . 



16 Fifth Avenue, New York. U. S. A. 

■and, Lopdo -land. 

id. , 

2, Berlin X. W., Germany. 

i)ore. Straits Settlements. 



CONTENTS 

&RILL, HARVEY C, and WELLS, ALBERT H. The physiological 

active constituents of certain Philippine medicinal plants: II.. 167 

BRILL, HARVEY C. The antineuritie properties of the infuso- 
rial earth extract of the hydrolyzed extract of rice polishings.. 199 

BRILL, HARVEY C, and WILLIAMS, ROBERT R. The use of 
chaulmoogra oil as a specific for leprosv.. 



u. s. 

The "Philippine Journal of Science" is issued as follow- currency. 

Section A. Chemical and Geological Sciences and the Industries.. 92.00 

Section B. Tropical Medicine - 3.00 

Section C. Botany - 2.00 

Section I). General Biology, Ethnology, and Anthropology (Sec- 
tion 1) began with Volume V) 2.00 

Entire Journal, Volume II, III, IV, or V ... 5.00 

Entire Journal, beginning with Volume VI .:.. 7.00 

Single numbers (except of Volume I) - 50 

Each section is separately paged and indexed. 
Authors receive lOO copies of their papers free. 
Volume I, 1906 (not divided into sections) and supplement, sold 

only with a complete file of section A, B, or C 10.00 

Supplement to Volume I <botany) 3,50 

Volume I (without supplement) , sold only with a complete file of 

section A, B, or C 6.50 

Single numbers ol Volume I ... .75 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, P. I., or to any of the agents 
listed below: 

AGENTS 

The Macmillan Company, 64-6C Fifth Avenue, New York City, TT. S. A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martlnus Nijhoff. Lange Voorhout 9, The Hague, Holland, 

Mayer & Miiller, Prinz Louis Ferdinandstrasse 2, Berlin, N. W., Germany. 

Kelly & Walsh, Limited, 3 2 Baffles Place, Singapore, Straits Settlements. 

A. M. & J. Fergnson, 19 Baillie Street, Colombo, Ceylon. 

Thackcr, Spink <& Co,, P. O. Box 54, Calcutta, India. 




-■■; '.■■■. . .'•■.':■■' . 



Vol. XII, Sec. A, No. 5 



September, 1917 



THE PHILIPPINE 

JOURNAL OF SCIENCE 



ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D.; J. R. WRIGHT, Ph. D,; G. W. HEISE, M. 
J. C. WITT, Ph. D.; T. DAR JUAN, A. B.; A. H. WELLS, A. B. 

r. c. McGregor, a. b.; h. e. kupfer, a. b. 




MANILA 
BUREAU OF PRINTING 



PUBLICATIONS FOR SALE BY THE BUREAU OE SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



ETHNOLOGY 

A VOCABULARY OF THE IGOROT IAN- 

GUAGE AS SPOKEN BY THE 

BONTOC IGOEOTS 

By Walter Clayton CLAPr 

Order No. 408. Paper, 89 pages, $0.75, 
postpaid. 
The vocabulary is given In Igorot-English 
and Enfllish-lflorot. 



THE NABALOI DIALECT 

By Otto Scueereb 

and 

THE BATAK8 OF PALAWAN 

By Edward Y. Mru.BR 

Order No. 403. Paper, $0.23: half mo- 
rocco, $0.75; postpaid. 
The Nabalol Dialect (65 pages, 29 
plates) and the Bataks of Palawan (7 
pages, 6 plates) are bound under one cover. 



THE BATAN DIALECT AS A MEMBER 

OF THE PHILIPPINE GROUP 

OF LANGUAGES 

By Otto Scheerbb 

and 

"F" AND "V" IN PHILIPPINE 
LANGUAGES 

By Carlos Everett Conant 

Order No. 407. 

These two papers are Issued under one 
cover, 141 pages, paper, $0.80, postpaid. 



ETHNOLOGY— Continued 



By Najeeb M. Saleeby 

Order No. 405. Paper, 107 pages, 16 
plates, 5 diagrams, $0.25; half mo- 
rocco, $0.75; postpaid. 
This volume deals with the earliest 
written records of the Mores in Mindanao. 
The names of the rulers of Maglndanao are 
recorded In five folding diagrams. 



NEGRITOS OF ZAMBALES 

By William Allan Reed 

Ordor No. 402, Papor, 83 pages, 62 
plates, $0.25; half moroooo, $0.75; 
postpaid. 
Plates from photographs, many of whloh 
were taken for this publication, show orna- 
ments, houses, men making fire with bamboo, 
bows and arrows, dances, and various typw 
of the people themselves. 



INDUSTRIES 

PHILIPPINE HATS 

By C. B. Robinson 

Order No. 415. Paper, 66 page*. 8 
Plates, $0.50 postpaid. 

This paper Is a conolse record ©f the 
history and present condition of hat making 
In the Philippine Islands. 



THE 8UBANUNS OF SINDANGAN BAT 

By Emerson B. Christus 

Order No. 410. Paper, 121 pages, 1 
map, 29 plates, $1.25, postpaid. 

Sindangan Bay is situated on the north- 
ern coast of Zamboanga Peninsula. The Su- 
banuns of this region were studied by Mr. 
Christie durJng two periods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanuns at 
work and at play; their industries, houses, 
altars, and . implements; and the people 
themselves. 



By Herbert S. Walker 

Order No. 412. Paper, 145 pages, 10 

plates, 1 map, $1.25, postpaid. 

Considered from the viewpoint of prac- 
tical utility, Mr. Walker's Sugar Industry 
In the Island of Nsgros Is one of the most 
Important papers published by the Bureau 
of Scienoe. This volume Is a real contribu- 
tion to the subject; it Is not a mere com- 
pilation, for the author was In the field and 
understands the conditions of which he 
writes. 



THE HISTORY OF SULU 

By Najeeb M. Saleeby 

Order No. 406. Paper, 275 pages, 4 

maps, 2 diagrams, $0.75, postpaid. 
In the preparation ef his manuscript for 
The History of Sulu, Doctor Saleeby spent 
much time and effort In gaining access 
to doouments in the possession of the Sultan 
of Sulu. This book is a history of the 
Moros in the Philippines from the earliest 
timts to the American occu ation. 



By Charles S. Banks 

Order No. 413. Paper, 53 pages, 20 
plates, $0.75, postpaid. 
In A Manual of Philippine Silk Culture 
are presented the results of several years' 
actual work with sllk-produoing larva to- 
gether with a description of the new Philip- 
pine raoe. 



THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 

Vol. XII SEPTEMBER, 1917 



THE COMPOSITION AND MOISTURE CONTENT OF THE SOILS 

IN THE TYPES OF VEGETATION AT DIFFERENT 

ELEVATIONS ON MOUNT MAQUILING 1 

By William H. Brown and Angel S. Arguelles 

(From the College of Liberal Arts, University of the Philippines, 
and the Bureau of Science, Manila) 

THREE PLATES AND ONE TEXT FIGURE 

In the Philippines, as in other moist tropical countries, there 
are frequently decided changes both in the composition and in 
the physical character of the vegetation even in very limited 
areas. One of the most remarkable examples of this is the 
frequency with which vegetation becomes dwarfed at compara- 
tively low elevation. On Mount Maquiling a lofty forest occurs 
at elevations up to 600 meters, while at an altitude of 1,000 
meters the ground is covered by elfin wood. The object of the 
work here reported was to determine whether or not the types 
of vegetation on the eastern and southeastern slopes of Mount 
Maquiling could be correlated with the composition and water 
content of the soils. 

The soil of the Philippines is largely of volcanic origin. Cox 2 
has published the results of a large number of analyses of soil 
from different parts of the Archipelago. He concluded that both 
the chemical and physical analyses indicated that the soil of the 
Islands was on the whole fertile. Cox also shows that the distri- 
bution of cultivated crops appears to be determined largely by the 

1 Received for publication June, 1917. 

' Cox, Alvin J., Philippine soils and some of the factors which influence 
them, This Journal, Sec. A (1911), 6, 279-330, Pis. I-XI. 

160586 221 




222 The Philippine Journal of Science isn 

character of the rainfall, certain plants being grown more ex- 
tensively where there is a distinct dry season and others where 
the rain is more evenly distributed throughout the year. A more 
intensive study of the soils of Luzon has been published by Cox 
and Argiielles. 3 

Brown and Matthews * have shown that in the Philippines the 
present distribution of forest and grassland is due to the activity 
of man combined with climatic influences. Grasslands only occur 
where the forests have been removed. They are produced and 
maintained by frequent fires and so are most extensive where 
there is a pronounced dry season. 

MOUNT MAQUILING 

Mount Maquiling is an isolated volcanic cone, situated on 
Luzon, midway between the eastern and western coasts, about 
64 kilometers southeast of Manila, in latitude 14° 10' east of 
Greenwich. It reaches an altitude of approximately 1,100 me- 
ters. The geology of the mountain, particularly of the lower 
slopes, has been described in considerable detail by Abella. s 

The main crater of the volcano has apparently been extinct 
for a long period. The rim has been eroded until it has become 
a series of peaks which, except on very steep slopes, are covered 
with deep soil. Volcanic activity has, however, not entirely 
ceased, as numerous fumeroles and hot springs occur around 
the base and on the lower slopes. 

On the eastern and southeastern sides the mountain grades into 
the surrounding plain at an elevation of approximately 50 meters. 
On the lower slopes there are layers of soft volcanic tuff. At 
elevations of from 50 to 100 meters these layers of tuff, which 
are mixed with layers of soil, may be many feet thick and are 
frequently near the surface. The layers near the surface are, 
however, usually thin, soft, and very much broken. Such a 
layer, about 30 centimeters thick, is frequently found about 30 
centimeters below the surface, and traces of such a layer occur at 
elevations as great as 300 meters or more. West and Cox 6 give 

* Cox, A. J., and Argiielles, A. S., The soils of the Island of Luzon, 
ibid., Sec. A (1914), 9, 1-50. 

* Brown, W. H., and Matthews, D. M., Philippine dipterocarp forests, 
ibid., .Sec. A (1914), 9, 413-561, Pis. I-XIII. 

1 Abella y Casariego, D. E., El Monte Maquilin (Filipinas) y sus ac- 
tuates unanaciones volcanicas. Iraprenta y Fundacion de M. Tullo, Madrid 
(1885), 1-28, Pis. 1-2. 

* West, A. P., and Cox, A. J., Burning tests of Philippine Portland 
cement raw materials, This Journal, Sec. A (1914), 9, 79. 



xii, a, 5 Brown and Argiielles: Soils on Mount Maquiling 223 

analyses of a number of samples of similar tuff. These analyses 
indicate that the tuff should disintegrate into a fertile soil. 
Where the surface layers of the tuff are shallow and much 
broken, they appear to have little if any effect in determining 
the character of the vegetation. We have seen thick surface 
layers only around the base, where the original vegetation has 
been almost entirely removed. 

The climate of Mount Maquiling may be classified as mon- 
soonal ; that is, the rains depend upon rain-bearing winds, which 
shift their direction twice a year. This results in distinct wet 
and dry seasons. The northeastern monsoon strikes Luzon on 
its eastern coast and deposits a large part of its moisture in 
passing over the divide between the Pacific Ocean and Laguna 
de Bay; when it reaches Mount Maquiling, it is a drying wind. 
This monsoon results in a decided dry season from January to 
April. The most pronounced rainy season is from July to Sep- 
tember during the southwest monsoon, when a large part of the 
rains are the result of cyclonic disturbances (typhoons). 

Around the base and on the lower slopes of Mount Maquiling 
there is a mixture of grassland and second-growth forest. Above 
this there are three distinct types of original forest, which occur 
at successively higher levels. 

GRASS AREA 

The mixture of grassland and second-growth forest at the 
base has been described by Brown and Matthews. 7 This region 
appears to have been originally covered with a tall dipterocarp 
forest, which is the type occurring at the next higher elevations. 
The original forest was removed, and the land was cultivated. In 
the area under consideration cultivation was abandoned, and 
much of the ground became covered by tall grasses, chiefly Sac- 
charum spontaneum (talahib) and Imperata exaltata (cogon). 
These grasses are very inflammable when dry. They were burned 
at frequent intervals, the last fire occurring in 1911. The grass 
fives kill nearly all tree seedlings, but appear to do little if any 
damage to the rhizomes of the grasses. Saccharum spontaneum 
(Plate 1, fig. 1) and Imperata exaltata both form dense stands, 
the former frequently reaching a height of over 3 meters, while 
the latter is shorter, being rarely more than 1.5 meters in 
height. Mixed with the grass are areas of second-growth 
forest, and since 1911 much of the grassland has changed to 

7 Op. cit. 



224 The Philippine Journal of Science 1917 

forest, as quick-growing trees readily invade the grass areas 
when there are no fires. The trees have a very rapid rate of 
growth and are small, usually reaching a height of only 10 
meters or less. 

Our soil samples from this altitude were taken in a grass 
area at an elevation of approximately 100 meters. 

DIPTEROCARP FOREST 

At altitudes between 200 and 600 meters the ground is covered 
by a tall, dense dipterocarp forest. This is composed of three 
distinct stories of trees, the tallest of which reaches a height 
of from 35 to 40 meters (Plate I, fig. 2). The most prominent 
tree in it is Parashorea plicata, which is frequently more than a 
meter in diameter. The second, or middle story, is composed of 
medium-sized trees, which spread their leaves under the branches 
of those of the top story. The second story reaches a height 
of about 18 or 20 meters. The most prominent species is 
Diplodiscus paniculatus (balobo), which is represented by many 
more trees than any other species in the forest and is probably 
about four times as numerous as any other second-story species. 
The third story is composed of small trees that reach a height 
of about 10 meters. 

The presence of the different stories is not evident on casual 
observation, as the specific composition of the stories is very 
complex and few of the trees present any striking peculiarities, 
while smaller trees of a higher story always occur in a lower 
story and between the different stories. In addition to the above 
stories of trees there is a ground cover composed largely of seed- 
lings of tree species and climbing palms (rattans), but also con- 
taining numerous herbs and shrubs. 

The foliage is so dense that when one walks through the forest 
most of the large trees are completely hidden from view. Plate 
II, fig. 1, shows an area from which the undergrowth and small 
trees have been removed. It will be seen that large trees are 
much more numerous than would be suspected from an examina- 
tion of Plate I, fig. 2, which shows a virgin area where there are 
probably as many trees as in the area shown in Plate II, fig. 1. 

For a fuller discussion of this forest, see an article by Brown 
and Matthews. 8 

The soil samples in the dipterocarp forest were obtained near 
the top of a ridge at an altitude of about 300 meters. 

$ Op. cit. 



xii, a, 5 Brown and Argiielles: Soils on Mount Maquiling 225 

MIDMOUNTAIN FOREST 

The forest occurring above the dipterocarp forest is much 
smaller than the latter and is composed of only two stories of 
trees (Plate II, fig. 2). This forest extends upward to an 
elevation of approximately 900 meters. The soil samples con- 
sidered in this paper were obtained on the top of a broad ridge 
at an elevation of about 730 meters. In this locality the trees 
of the top story reach a height of about 18 meters. The forest 
is much more open than the dipterocarp forest (Plate III, fig. 1). 
The most prominent large tree in it is Quercus solariana (cata- 
ban), while Cratoxylon celebicum (guyong guyong), which is 
somewhat smaller, is more numerous. The average height of 
the second story is about 6 or 8 meters. The most numerous 
species are Oreocnide trinervis (malatuba) , Neolitsia villosa, and 
Sauraunia barnsii. The ground cover consists largely of ferns 
and other herbs, of which species of Elatostema are the most 
prominent. The herbs in this region require much moister 
conditions than those in the dipterocarp forest. 

MOSSY FOREST 

The top of the mountain is in the cloud belt and is covered 
with an elfin wood or mossy forest. This type is composed of 
only one story and is characterized by having the trunks of the 
trees thickly covered by mosses, mosslike plants, and other epi- 
phytes. The ground cover is composed almost entirely of herbs, 
among which Strobilanthus plurifomis, ferns, and species of 
Selaginella are the most prominent. The most striking peculiar- 
ity of the trees is a tendency to produce a large number of aerial 
roots, which grow to such a size as to form, as far as function 
goes, secondary trunks (Plate III, fig. 1). The soil samples 
from this forest were taken within a few meters of the top of 
the mountain. 

From the above description it will be seen that the chief 
changes caused by increased altitude, which might conceivably 
be connected with soil conditions, are a dwarfing of the vegetation 
as higher elevations are reached and the occurrence in the ground 
cover of plants requiring moister conditions at high than 
at low altitudes. In order to see whether or not these changes 
could be connected with soil conditions, we have made a chemical 
and physical analysis of soil samples taken at a depth of 20 
centimeters in the four different types of vegetation and have 
determined the water content of the soils, for different weeks 



226 



The Philippine Journal of Science 



of the year, in the same locations at depths of 10, 20, and 30 
centimeters. 

CHEMICAL COMPOSITION 

In Table I are given chemical analyses of the soils from the 
different altitudes. 

Table I. — Chemical analyses of soils from Mount Maquiling, Laguna 
Province, Luzon. 

[Water-free basis, numbers give percentages.] 





Source of soil. 


Grass- 
land. 


Diptero- 
carp 
forest. 


Mid- 
moun- ' 

tain 
forest. 


Mossy 
forest. 




10.32 
0.150 
0.278 
0.86 
0.62 
0.294 
0.37 
1.36 


10.08 
0.137 
0.106 
1.01 
0.61 
0.241 
0.44 
1.06 


13.56 
0.199 
0.104 
0.31 
0.49 
0.189 
0.53 
1.71 
0.0094 


29.97 
0.644 
0.112 
0.52 
0.79 
0.170 
0.34 
8.06 
0.0082 






Lime (CaO) 

















It will be seen that none of the soils are strikingly deficient in 
any important element and none can be considered as acid. The 
amount of nitrogen and humus is greatest in the mossy forest, 
where the vegetation is most dwarfed. It is to be noted that 
the smallest amount of nitrogen and humus is shown by the 
sample from the tall dipterocarp forest. The chemical analyses 
of the soil, as shown in Table I, do not indicate that there is any 
connection between the chemical composition of the soil and the 
dwarfing of the vegetation as higher elevations are reached. 

PHYSICAL COMPOSITION 

The physical analyses of the soils from the different types of 
vegetation are given in Table II. 

There is nothing in these analyses to indicate that all of the 
soils should not produce a luxuriant type of vegetation. If any 
distinction can be made, it seems that the soil of the mossy 
forest should be the best, as this is a fine sandy loam, while the 
soil of the dipterocarp forest and grassland is a loamy clay. 



xii. a. 5 Brown and Argiielles: Soils on Mount Maquiling 227 

Table II. — Mechanical analyses of soils from Mount Maquiling, Laguna 
Province, Luzon. 

[Water-free basis, numbers (five percentages.] 



Classification of soil. 


Diameter of particles. 


Source of soil. 


Grass- 
land. 


Dip- 

ter- 

ocarp 

forest. 


Mid- 
moun- 
tain 
forest. 

nil 
0.7 
0.9 
4.1 
8.8 
12.3 
4.9 
39.4 
29.6 


Mossy 
forest. 






nil 
0.9 
2.2 
5.4 
12.5 


nil 
0.5 
1.3 
4.5 
9.8 


nil 
nil 
5.9 
23.3 
31.4 
13.2 
11.2 
9.4 
5.6 


















10.8 1 7.8 
13.6 | 13.5 
36.5 41.7 
19.0 21.4 























MOISTURE CONTENT 
In Tables III to VI is shown the percentage of moisture in 
the soil at the different altitudes. 

Table III. — Percentage of soil moisture in grassland (altitude 100 meters) 
at base of Mount Maquiling, Laguna Province, Luzon. 



1912. 
December 6 ... 

13 ... 
20... 
27.. 
1913. 

January 3 ... 

10... 

17... 

24 ... 

31.. 
February 7 .. 

14 .. 
21... 
28 ... 

March 7 ... 

14 ... 

21 .. 

28.. 
April 4 ... 

11.. 

18- 



34.5 

34.7 
34.6 



39.4 
35.5 

36.4 
33.4 
30.4 
34.6 
33.6 
34.6 



24.2 
26. 5 
30.6 



28.3 
28.3 
28.6 
27.7 
27.1 



1913-Cont. 

May 2.. 

9 

16 

23 

30 

June 6 

20 

July 18 

25 

August 1 

15 

22 

29 

September 5 

26. 

October 3 

24 

November 7 

14 

21 



34.3 

85. 4 
38.7 
37.1 



27.8 
27.1 
30.3 
32.8 
33.1 
31.9 
29.9 
35.1 
36.0 
35.7 

34.7 
36.8 
35.2 
40.1 
40.6 
36.9 
33.4 
31.7 
34.3 
42.6 



228 



The Philippine Journal of Science 



One of the most striking peculiarities about these tables is 
the similarity in the amount of moisture in any given soil at 
different depths. This is apparently due to the fact that the 
dense covering of the vegetation prevents the surface soil from 
drying out at a much more rapid rate than the deeper layers. 

The average percentage by months of the moisture in the soil 
at a depth of 20 centimeters, at elevations of 90, 350, and 725 
meters, is shown graphically in fig. 1. 




Fig. 1. Average percentage, by months, of moisture in the soil at a depth of 20 centi- 
meters at different elevations on Mount Maquiling. 

The lowest moisture content is shown by the grassland at the 
base of Mount Maquiling (Table III). Here, however, the 
growth of trees is much more rapid than at any of the higher 
elevations. This is very probably not due at all to soil con- 
ditions, but to the greater illumination of the individual trees 
and the rapid rate of growth characteristic of the second-growth 
species. 

At a depth of 10 centimeters the greatest variation of water 
content is from 41.3 per cent on November 21 to 24.1 per cent 
on April 11. The soil samples taken in different places in a 
very limited area on the same day would, of course, show varia- 



xii, a, 5 Brown and Argiielles: Soils on Mount Maquiling 229 



tions in water content. Some of the variations seen in Table III 
are undoubtedly due to the selection of individual samples and 
are not connected with any change in the water content of the 
soil as a whole. It is probable that this source of error is 
responsible for the extreme variations noted in the water con- 
tent given in Table III. The soil at a depth of 10 centimeters 
gives an average water content for November of 38 per cent 
and for April of 27 per cent. These figures would probably 
represent the extreme variations for the region as a whole more 
accurately than the highest and lowest figure shown in Table III. 
The figures for April show that even in the dry season the soil 
contains considerable moisture. The lower moisture content 
combined with the dry atmospheric conditions at this season are, 
however, sufficient to affect the plants very adversely. Many of 
the second-growth tree species loose their leaves completely for 
longer or shorter periods, and all of them show a much lower rate 
of growth during the dry season than at other times of the year. 

Table IV. — Percentage of moisture in soil in dipterocarp forest (altitude 
300 meters) on Mount Maquiling, Laguna Province. 



1912. 
November 15 ... 

22... 

29... 
December 6 ... 

13 — 
20... 

27 — 
1913. 

January 3 ... 
10... 
17 — 
24 — 
31... 

February 7 ... 
14... 
21... 

March 7 ... 

14 ... 
21 ... 

28 ... 
April 4 ... 



Depth in centi- 
meters. 

i 


10. 


20. 


30. 


57.4 


57.1 


1 
56.6 j 


57.2 


59.5 


56.5 


57.2 


57.8 


58.6 | 


55.2 


57.8 


56.5 


53.1 


54.4 


52.3 


54.2 


52.4 


56.2 ! 


60.4 


54.6 


54.5 


65.5 


54.7 


57.0 


53.8 


52.2 


54.2 ! 


51.2 


47.2 


54.9 1 


53.6 


57.4 


60.0 


54.6 


54.4 


56.3 1 


47.1 


45.8 


51.7 


47.6 


50.5 


53.5 


55.6 


52.5 


48.0 j 


44.9 


48.1 


47.1 


47.5 


52.4 


51.5 


45.9 


46.7 


47.9 


39.0 


45.0 


50.3 


40.6 


44.1 


48.8 


42.4 


41.4 


44.5 


41.3 


42.6 


44.5 


52.1 


49.4 


49.8 


42.1 


47.6 


44.8 



1913-Cont. 
May 2 



July 

August 



September 5. 

12. 

19. 
October 3. 

24. 
November 7. 

14. 



December 12. 



54.8 
53.0 
58.1 
64.5 
55.4 
54.8 
57.0 
55.4 
56.1 
55.2 
55.1 

55.2 
57.2 



[ 55.4 
59.1 



230 



The Philippine Journal of Science 



The soil of the dipterocarp forest contains considerably more 
water than that of the grassland. The greatest variation in 
moisture content at a depth of 10 centimeters is from 59.1 per 
cent on December 19 to 39 per cent on March 21. The latter 
figure appears to be high for a minimum water content at this 
depth, but during the dry season the effects of drought on the 
dipterocarp forest are very evident. The foliage is much less 
dense than at other seasons, while Parastaca plicata (bagtican 
lauan) , the dominant tree, shows a greatly diminished rate of 
growth. 9 The effect of drought in the dipterocarp forest is, 
however, not so marked as in the grassland. 

The percentages of moisture shown in Table V for the mid- 
mountain forest are considerably greater than those for the 
dipterocarp forest, and the variation is relatively less. 

Table V. — Percentage of moisture in soil in midmountain forest (altitude 
730 meters) on Mount Maquiling, Laguna Province, Luzon. 



1912. 
November 16 



30. 

December 7. 

14. 

21. 



1913. 
January 11 



77. 2 j 80. 1 

| 75.6 

76.8 | 

70.2 I 73.6 
69.1 ! 72.8 



73.7 69.7 

74.6 I 73.1 

70.1 | 70.7 
57.2 



1913— Cont. 

May 3 

10 



July 
August 



October 4. 

25. 

November 1. 

29_ 

December 6. 

13. 



In this region there were no marked effects of droughts during 
the dry season except on the epiphitic vegetation. The percent- 
age of moisture in the soil is very high, being between 70 and 80 



" Brown and Matthews, op. cit. 



xii. a. 5 Broivn and Argiielles: Soils on Mount Maquiling 231 



per cent for a large portion of the year. There is, however, 
no indication that this amount is excessive, as the soil is well 
drained. 

Table VI. — Percentage of moisture in soil in mossy forest (altitude 1,070 
?ncters) at top of Mount Maquiling, Laguna Province, Luzon. 



1912. 

November 16 

23.— 



14... 

21.. 

1913. 

January 11 .. 



Depth in centimeters. 


10. 


20. 


30. ; 


258.1 


190.5 


j 

246.7 I 


251.5 


167.5 


121.1 


305.0 


312.4 


129.5 


275.9 


258.2 


123.2 i 


291.0 


188.1 


180.5 


234.5 


352.8 


232.0 


245.0 


268.2 


278.0 


178.6 


149.5 


362.3 


130.6 


234.3 


305.8 


211. 1 


225.2 


289.0 


250.5 


188.il 


206.0 | 




320.2 


319.9 


281.8 


256.0 


255.5 


257.2 


198. S 


198.6 


221.2 


264.9 


232.1 


203.6 


221.8 


257.2 


306.3 


114.6 




167.4 


88.2 


75.1 


176.5 


152.0 


139.2 



July 

August 



September 27. 
October 4. 

18. 

25. 
November 8. 



Depth in centimeters. 



237.7 
273.8 



218.3 
238.0 
218.8 
148.1 
159.9 
185.7 
154.2 
247.3 
165.3 
134.6 



136.8 | 122.0 

214.5 173.6 

185.4 I 173.6 

246.7 251.8 

161.7 I 162.4 



171.0 . 
137.2 I 



211.8 
207.6 



257.1 ; 

. ; 173.1 

164.2 j 256.2 

305.5 | 247.9 

186.1 | 302.0 

263.2 i 301.2 

203.6 | 256.5 
204.8 j 272.1 



The moisture content of the soil in the mossy forest (Table VI) 
is extremely high. There is only one week when the determina- 
tions show less than 100 per cent of moisture, and on this occasion 
this is true of only two of the three depths. Equally as striking 
as the high moisture content is the variation from week to week 
and at different depths on the same week. The different depths 
frequently show a variation of more than 100 per cent on the 
same day. The high moisture content is apparently connected 
with the large amount of organic matter found in this soil. The 
amount of organic matter apparently varies greatly in different 
situations, and the variations in the water content, shown in 
Table VI, are probably due more to the places in which the 
samples were taken than to any weekly change in the moisture 
content of the soil as a whole. 

The high moisture content shown by the soil in the mossy 
and the midmountain forests undoubtedly accounts for the fact 



232 The Philippine Journal of Science 

that the ground covering is composed of plants requiring more 
moisture than those found in the dipterocarp forest. 

It does not seem probable, however, that the dwarfing of the 
vegetation is connected with this high water content. The 
great amount of water in the soil of the mossy forest is not due 
to heavy rains that would cause the soil to become leached out, 
as the rainfall here is about the same as it is in the place where 
the soil samples were taken in the dipterocarp forest and is con- 
siderably less than in the midmountain forest and in a large pro- 
portion of the dipterocarp forest. The soil of the mossy forest 
is, moreover, well drained, so that it must be well aerated, and 
it has a springy consistency. The figures in Table I show less 
acidity in the soil of the mossy forest than in that of the dip- 
terocarp forest. There seems, therefore, to be no reason for 
considering the high moisture content of the soil as harmful. 
The moisture content of the soil of the midmountain forest would 
certainly not seem to be high enough to be deleterious to the 
vegetation. The fact that the trees in this situation are much 
smaller than those in the dipterocarp forest is probably due to 
the same factors that have resulted in the stunted vegetation 
on the top of the mountain, the difference being that these factors 
are more pronounced at the top. 

The above discussion indicates that the moisture contents of 
the soil should be as favorable at high as at low altitudes and 
so cannot be connected with the dwarfing of the vegetation as 
higher altitudes are reached. 

SUMMARY 

The natural vegetation of Mount Maquiling becomes more and 
more dwarfed as higher elevations are reached. There does not 
appear to be anything in the physical or mechanical composition 
of the soil or its moisture content which would account for this 
fact. 

The character of the plants in the ground cover varies accord- 
ing to the amount of moisture in the soil at different altitudes. 



ILLUSTRATIONS 

Plate I 

Fig. 1. Growth of Saccharum epontaneum, at an altitude of about 90 meters 
on Mount Maquiling. Photograph by Brown. 
2. View in dipterocarp forest, Mount Maquiling, at an elevation of 
about 500 meters. The large tree in the center is an individual 
of Parashorea plicata in the first story, the smaller trees on the 
right are in the second story, while still smaller third-story 
species are scattered throughout the picture. The feathery leaves 
of the climbing palms (rattans) are the most conspicuous elements 
in the undergrowth. The density of the vegetation is very 
evident, the foliage of the undergrowth and lower stories being 
so dense that most of the large trees are completely hidden. 
Photograph by Brown. 

Plate II 

Fig. 1. View in dipterocarp forest, Mount Maquiling, altitude about 300 
meters. The undergrowth and all small trees have been removed. 
The clearing was done by the College of Agriculture for the 
purpose of planting coffee, and some of the trees removed were 
as much as a meter in diameter. In the forest shown in Plate 
I, fig. 2, there are probably as many trees as in the one shown in 
this picture, the difference in the number of trees seen in the 
pictures being due to the fact that in the former case the trees 
are hidden by the foliage, while in the latter they are in plainer 
view. Photograph by Brown. 
2. View along a trail in the midmountain forest, Mount Maquiling, 
at an altitude of 740 meters. The vines that are prominent, 
particularly in the left of the picture, are species of Freycinetia. 
Fruits can be seen growing on the trunk of the large Ficus to 
the right. A comparison of this view with Plate I, fig. 2, will 
show that the midmountain forest is much more open than the 
dipterocarp forest. Photograph by Brown. 

Plate III 

Fig. 1. View in midmountain forest, Mount Maquiling, altitude about 730 
meters. This view shows the undergrowth and the open charac- 
ter of the midmountain forest even better than Plate II, fig. 2. 
Photograph by Brown. 
2. Large aerial roots of a tree in the mossy forest on Mount Maquiling. 
Photograph by Brown. 

text figure 

Fig. 1. Average percentage, by months, of moisture in the soil at a depth 
of 20 centimeters at different elevations on Mount Maquiling. 

233 



Broun and AboOelles: Soils <>n Mount Maquii.inh.1 fPni!,. Journ. Sii., XII, A, No. 5. 




Fig. 1. Saccharum spontaneum on Mount Maquiling: altitude. 90 meter 



----- ■./-'■ ;- 
'-' -- ■ - - : - '■■"' -': fit ■'-.'."■ ■" 


1 

*r ■■_• v 






PfJl 












;-;^4;X:>a 


; 


fe- : 


ShH^j^^ 


•:% 






: 






■HE^^^E 



Fig. 2. Dipterocarp forest, Mount Maquiling; elevation, 500 meters. 
PLATE I. 



Brown ami Arci ki i.i:s : Soils on Moi'nt Maiji'iunc] ITiiil. Journ. Sci., XII. A, No. 5. 




Fig. 1. Dipterocarp forest with undergrowth and small trees removed. Mount Maqu 
elevation. 300 meters. 




Fig. 2. Trail in midmountain fohest, Mount Maquiling: elevation, 740 meters. 
PLATE II. 



Brown and Argi i .i.i.fs : Soils on Mount Maqi-iling.] [Phil. Jourx. Sci., XII, A, No. 5. 




Fig. 1. Midmountain forest. Mount Maquiling: elevation, 730 meters. 




A COMPARISON OF LINSEED OIL AND LUMBANG OILS AS 
PAINT VEHICLES " 

By R. H. Aguilar 

(From the Laboratory of General, Inorganic, and Physical Chemistry, 
Bureau of Science, Manila) 

ONE PLATE AND ONE TEXT FIGURE 

Because of its general adaptability to different kinds of paints, 
linseed oil has not been supplanted 2 on a commercial scale by 
any other oil. The lumbang oils are possible substitutes for 
linseed oil, and they have been much studied, 3 but little has been 
reported concerning their behavior with different pigments or the 
quality of the resulting paints. It is, therefore, hoped that the 
following preliminary series of comparative tests on the proper- 
ties of linseed, lumbang bato, and lumbang banucalag oils may 
be of interest. 

Lumbang bato (Aleurites moluccana), a large tree belonging 
to the family Euphorbiacea?, is common and widely distributed in 
the Philippines, occurring in most islands and provinces. It 
occurs both as a native and as a semicultivated tree and is locally 
abundant. The species is one of very wide geographic distribu- 
tion, extending from India through Malaya to Polynesia. It is 
commonly known as the candlenut tree, but has numerous local 
names in the various countries where it occurs. In Hawaii it 
is known as kukui. 

According to Richmond and Rosario 4 the seed yields from 60 
to 65 per cent oil by extraction with carbon bisulphide, ether, or 
chloroform and 55 per cent by hydraulic expression at 500 kilo- 
grams per square centimeter. In the present work the yield 
of oil was 44 per cent (calculated on the weight of the kernels) 
by hydraulic expression at 310 kilograms per square centimeter. 

1 Received for publication July, 1917. 

2 Gardner, H. A., Paint Technology and Tests. McGraw-Hill Book Co., 
New York (1911), 2. 

3 Wilcox, E. V., and Thomson, A. R., Bull. Hawaii Agr. Exp. Sta. (1913), 
39. Brill, H. C., and Agcaoili, F., This Journal, Sec. A (1915), 10, 105-119. 

4 Richmond, C. F., and Rosario, M. V., ibid., Sec. A (1907), 2, 441. 

235 



236 



The Philippine Journal of Science 



Lumbang banucalag (Aleurites trisperma) is confined to the 
Philippine Islands. Though of wide geographic distribution in 
the Archipelago, extending from central Luzon to Mindanao, it 
is less common and much more local than is lumbang bato. The 
oil is darker colored and more viscous than lumbang bato oil. It 
is commonly supposed to cause skin eruptions on contact, but I 
have seen no foundation for this belief. I have handled the oil 
constantly in this work, but have suffered no inconvenience. 
The yield, calculated on the kernel weight, is about 43 per cent by 
hydraulic expression at 310 kilograms per square centimeter. 

Brill and Agcaoili 5 found that the lumbang oils are comparable 
with linseed oil in drying quality of film and percentage change 
in weight when drying ; also that they can be used as substitutes 
for tung, or Chinese wood, oil, the product of allied species of 
the same genus, Aleurites fordii and Aleurites cor data. 

EXPERIMENTAL PART 

The oils used in this series of tests have the constants noted in 
Table I. 

Table I. — Oil constants. 1 







Fresh lumbang:. 


seed oil. 


Bato. 


Banucalag:. 

0.9368 
197. 74 
145.25 

8.70 




0. 9345 
189.40 
183.96 

0.50 


0.9252 
192.95 
150.36 

1.05 











" Analyzed by F. Agcaoili, chemist, Bureau of Science. 

All the tests included in this preliminary work have been 
carried on under ordinary Philippine laboratory conditions, that 
is to say, in diffused daylight, at a temperature of from 28° to 
30° C, and at a relative humidity of about 75 per cent. 

Drying test. — The drying tests were made in the usual way 
by spreading a small amount of oil to uniform thickness on 
glass plates, each 6.35 by 8.89 centimeters (2.5 by 3.5 inches) 
in size. The oil films were placed in a glass case and were 
exposed to a dry atmosphere obtained by passing a slow current 
of air through sulphuric acid. The changes in weight and ap- 
pearance are shown in Table II and in fig. 1. 



6 Op. cit. 



xii, a, 6 Aguilar: Linseed and Lumbang Oils 237 

Table II. — Comparative drying tests with linseed and lumbang oils.' 



No. of 
curve. 


Oils used. 


Maxi- 
mum in- 
crease in 
weight. 


Day 
when 
maxi- 
mum in- 
crease in 
weight 
was 
attained. 


Condition of film. 


1 
2 
3 

4 

e 

6 


Linseed, boiled commercial sam- 
ple. 
Linseed, raw. of the best quality. 

Lumbang bato, bottled six 
months. 


Per cent. 
13.85 

12.13 

11.20 

11.03 

8.30 
8.92 


1 
4 
2 

4 

4 


Perfectly dry, clear, and firm in 

one day. 
Dry, clear, and firm between 

the fourth and the fifth day. 
Dry, clear, and firm between the 

second and the third day, tacky 

at the end of twenty days. 
Dry between the fourth and 

the sixth day. 
Dry and firm in two days. 
Dry in four days, but slightly 

opaque. 


Lumbang: banucalag, boiled 

Lumbang: banucalag. fresh 



a Average weight of oil taken was 0.1415 gram. 

The oil films were dry and firm about the time the maximum 
increase in weights was attained. 

Fig. 1 shows the similarity in the behavior of the three oils. 
The boiled and the aged oils dry much more rapidly than the 
fresh oils, and the curves of weight increase are almost straight 
lines from the origin to the maximum point. On the other hand, 
the fresh oils dry very slowly the first day, then more rapidly, 
until the maximum increase in weight is attained. 

Redman and others 6 give 11.7 per cent as the maximum in- 
crease in weight for linseed oil at the end of the sixth day and 
10.5 per cent for tung oil between the eighth and the ninth day. 
Lippert 7 gives 12.4 per cent as the maximum increase in weight 
for linseed oil. The lumbang oils are, therefore, similar in 
this respect to linseed and Chinese tung oil. 

Paint films. — The following methods for the preparation of 
paint films have been tested in the laboratory: 

1. The mercury method consists in allowing a paint to spread 
on top of mercury and later removing the dried paint film by 
lifting it off the liquid metal. Films obtained in this manner 

' Redman, L. V., Weith, A. J., Brock, F. P., Journ. Ind. & Eng. Chem. 
(1913), 5, 630. 

'Lippert, W., Zeitschr. f. angew. Chem. (1898), 412. 



238 



The Philippine Journal of Science 



are not uniform; they become thicker in the center than at the 
edges because of surface tension. 

2. The amalgam method consists in painting over a piece of 
tin plate. 8 After the paint is dry, it is scratched at a few 
points and drops of mercury are placed on the scratches. The 
mercury amalgamates with the tin, and the paint film can be 
lifted off the semiliquid amalgam. The quantity of tin and mer- 
cury needed, the necessity for using only unrusted tin plate, and 
the time required to remove the film from the tin militate against 
the usefulness of this method. 







, 












Time 


in days. 

10 


12 13 I 








IS I 


















































N 
























LEGEND 
















\ 


















— — — — Lmieed oil boiled. 

iumbang boto oil, 6 months old. 

Lumbang banucolog oil, boiled. 

— ■■■"■■ Linseed oil fresh. 

—iumbang bato oil, fresh. 

iumbona banucolog oil, fresh. 












V' 






V 










*; 




i 


Nl s 


^^ 


J\ 










% 




i 
i 






N > 
















i 








.5 




i 


Jr 


\ s 


'• 


\ 


"-- 






















1 




I f 


lit 


■•■-.. 


-•^ 


<nL 


\ 


«». 








=Ct ^ 




^tr— — 








60 




// 


I'l 








-•- 


^v. 




"*•- 















"""*■ 


T* 




jjj 




i ■ 


I 

i 














^ 


-•^ 


^ 


^ 


^ 




=^ 






= 


&< 5 


'1/ i 
h; III! 




















"*"• 


-•- 


t 


-- 




( / 


> 'i 


— 






































































ti 


! i J 
UA 






































1J\ 





































Fio. 1. Comparative drying tests for linseed, lumbang bato, and lumbang banucalag oils. 

3. Lipowitz or Wood's metal can be also used. The metal is 
coated with the paint under examination. After the paint is 
dry, the metal is put on a smooth plate and gently warmed. At 
a temperature below its melting point (60° C.) the metal softens 
sufficiently to allow the stripping of the paint. Satisfactory 
paint films can be made in this way. This process is not well 
adapted to laboratory work because of the large amount of ex- 
pensive alloy required and the care necessary in removing a film. 

' Emsmann, D. H., in Brannt, W. T., Varnishes, Lacquers, Printing Inks 
and Sealing Waxes. H. C. Baird Co., Phila. (1893), 321. 



xii. a, b Aguilar: Linseed and Lumbang Oils 239 

The most satisfactory results, however, were obtained by paint- 
ing over paper of fair quality, sized with some inert substance 
from which paint could readily be removed by soaking in warm 
water. Among the substances tried were library paste, agar- 
agar, sugar sirup, and glue. Of these, the use of paste or agar 
resulted in a shriveled film of low tensile strength. The use 
of sugar sirup 9 was satisfactory, but since the glue method, as 
described by Gardner, ,n was simpler and gave more uniform 
results, the latter was adopted for this work. The method of 
procedure was as follows: The paint was applied to the sized 
paper and allowed to dry. It was then removed by immersion 
in water at 40° C., washed off with fresh water to remove the 
glue, hung on a glass rod to dry, and placed in a suitable con- 
tainer to prevent the accumulation of dust and insects. 

Tensile strength and elasticity. — In testing the paint films, a 
modified Gardner-de Horvath n film-testing machine was em- 
ployed. The apparatus is shown on Plate I. 

The pressure was measured by means of a mercury manom- 
eter. This was made self-registering by placing inside the 
manometer tube an indicator, which remained at the highest 
point reached by the mercury column. The stretch was recorded 
by means of an aluminium rod resting on the center of the film. 
This rod was in turn fastened to the short arm of a lever and so 
delicately counterpoised that it exerted no appreciable pressure, 
yet rose and fell with any movement of the film. As the long, 
or pointer, arm of the lever was ten times as long as the short 
arm, it is obvious that any movement of the film was transmitted 
tenfold to the pointer arm and could thus be estimated with 
a certain degree of accuracy. 

To study the behavior of oil films, 12 precipitated silica passing 
through a 200-mesh sieve was incorporated with the oil in the 
proportion of 15 per cent silica by weight. The films were made 
of three coats of 2.5 grams of paint each, applied to 400 square 
centimeters of sized paper. 

Table III shows the effect of age upon the strength of films 
made from mixtures of linseed oil with lumbang bato oil in 
different proportions. 

9 Labordere, P., and Anstett, F., Chem Eng. (1913), 17, 1. 

10 Op. cit., p. 71. 

11 Gardner, op. cit., p. 79. 

12 Gardner, op. cit., p. 31. 



240 The Philippine Journal of Science 1917 

Table III. — Effect of age upon the strength of oil-silica films. 




The films after sixty days were tacky and soft and of greatly 
reduced tensile strength. 

To study the effect of age upon the strength of paint films 
prepared from active pigments, red lead was employed. The 
paints were prepared by incorporating 40 grams of red lead with 
25 grams of oil, and the films were the same in weight as the oil- 
silica films previously described. Table IV shows the relation 
between the increase of strength and the age of the paint films. 

Table IV. — Tensile strength of red lead paint films. 



I Breaking strength in terms of 
I centimeters of mercury. i 



Stretches in centimeters. 



Age of film in days. 



Linseed 1 

Do I 

Lumbang bato 

Do _ ... 



18.9 

16.1 

50,1 
100 10.7 



Age of film in days. 
. __ _ 

40. 



0. 22 I 0. 18 ' 0. 18 



0.26 ! 0.24 | 0.26 0.24 



Table IV shows that the paint films attained their maximum 
strength between the third and the fourth month and that then 
a gradual decrease followed. 

Further tests were conducted to compare the paint properties 
of lumbang banucalag with those of linseed oil and lumbang bato 
oil. Twenty-five grams of lumbang banucalag oil were mixed 
with 40 grams of red lead ; the mixture became thick and pasty in 
fifteen minutes, a property by means of which it can be differen- 
tiated from lumbang bato, because the latter, very much like 



Aguilar: Linseed and Lumbang Oils 



241 



linseed oil, does not show this phenomenon. The paint made 
with lumbang banucalag oil and red lead, painted on a smooth 
surface, did not dry in twelve days, probably because a very thin 
layer of the exposed surface dried in a very short time, thus 
preventing the action of the atmosphere upon the underlying 
paint. This property makes banucalag undesirable as a paint 
vehicle when used alone. However, by mixing 1 part of lum- 
bang banucalag with from 1 to 3 parts of lumbang bato, a quick- 
drying and altogether satisfactory paint is obtained. 

Table V shows the variation in the tensile strength of paint 
films made from mixtures of lumbang oils. 

Table V. — Tensile strength of red lead paint films. 









Breaking 

strength in terms 
of centimeters of 


Stretches in centi- 
meters, after— 




No. 


Oils used. 


Per 
cent. 


mercury, after — 




General remarks. 


10 
days. 


30 
days. 


60 
days. 


10 30 
days. days. 


60 
days. 


T 




100 


26.3 


34.0 


38.0 


1 
0.27 ! 0.26 ' 0.29 


















| 


hours. Appearance 
















good. 


2 


Lumbang bato._ 


100 


10.8 


17.4 


20.5 0.25 


0.28 


0.29 


Film was dry in 48 
hours. Appearance 


3 


[Lumbang bato 

1 Lumbang banucalag . 


90 
10 


] 10. 2 | 15.7 


15.5 


0.25 


0.24 


0.24 


good. 
Do. 


4 


J Lumbang bato 

1 Lumbang banucalag _ 


75 
25 


Jl2.4 17.6 


17.1 


0.32 


0.31 


0.26 


Film was dry in 24 
hours. Appearance 


5 


|Lumbang bato 

1 Lumbang banucalag . 


50 
50 


ll2.4 1 18.0 


18.4 | 0.30 


0.35 


0.33 


good. 
Do. 


6 


J Lumbang bato 

1 Lumbang banucalag . 


25 

75 


J-16.1 j 20.2 


21.2 


0.25 


0.22 


0.21 


Film was not dry in 6 
days. The surface 
was dull. Paint 
















dried into a paste in 




















the container in a 


7 




10 
90 


}l4.2 


« 


« 


0.24 


0.27 


0.25 


short time. 
Film was not dry in 9 
days. The surface 


1 Lumbang banucalag . 




















was also dull. Paint 




















dried into a paste in 




















the container in a 




















short time. 



The above results indicate that the mixtures of lumbang oils 
attain their maximum strength more quickly than either linseed 
or lumbang bato. Judging from the general behavior and ap- 
pearance of the paints, mixtures 4 and 5, that is, mixtures with 
lumbang banucalag containing 50 to 75 per cent lumbang bato, 



242 



The Philippine Journal of Science 



will be the most desirable for a red lead paint. With these 
mixtures, the resulting paint does not set as in the case of lin- 
seed u or lumbang bato oils ; it does not dry into a paste in the 
container as in the case of mixtures 6 and 7 ; and a better paint 
is obtained than can be secured with linseed or lumbang bato 
oils alone. 

Moisture-excluding property. — To determine the moisture- 
excluding property of films, 14 pieces of suitable size and shape 
were fastened with Canada balsam over the mouths of 200 cubic 
centimeter bottles containing drying agents. One set of bottles 
was half filled with concentrated sulphuric acid, another set 
with calcium chloride. The test pieces were cut from the same 
films as those whose tensile strength was recorded in Table IV. 
The moisture-excluding property of linseed and lumbang oils 
as determined by the increase in weight of the bottles of drying 
agent is shown in Table VI. 

Table VI. — Moisture experiment. 

[Numbers express percentage gain in weight. The calculations were based on the original 
weights of concentrated sulphuric acid and calcium chloride.] 



Oils used. 


Per 


Absorbent. 


Number of days. 


cent. 


5. 


10. 


40. 




100 
50 
50 

100 

100 
50 
50 

100 




0.24 
0.37 

0.42 
1.07 


0.45 
0.69 


1.67 
2.53 


Do 


f — -do 

CaCl2 


Lumbang 




2. CO | 6.98 


Do 


Lumbang 

Do 


(-;— " " 


1.14 
1.18 


2.15 
2.18 


7.52 
7.76 



Table VI shows that the moisture-excluding property of lum- 
bang bato red lead paint films are not as high as those of linseed 
oil red lead paint films. The moisture absorption with calcium 
chloride was relatively higher than with sulphuric acid. 

Similar tests were conducted with the films made from mixtures 
of lumbang bato and lumbang banucalag oils whose strength 
was shown in Table V. In this test 25 cubic centimeter bottles 
half filled with concentrated sulphuric acid were used. The 
results obtained are shown in Table VII. 



" Cf. Holley, C. D., Analysis of Paint and Varnish Products, 
and Sons, New York (1912), p. 222. 
" Gardner, op. cit., p. 83. 



John Wiley 



Aguilar: Linseed and Lumbang Oils. 
Table VII. — Moisture experiment. 

[Numbers express percentage gain in weight.] 



243 



No. 


Oils used. 


Per 
cent. 


Number of days 




6. 


12. 


32. 


60. 


1 
2 

4 
5 
6 

7 




100 
100 
90 
10 
75 
25 
50 
50 
25 
75 
10 
90 


0.05 

0.05 

I 0.07 

I 0.06 

1 0.06 

1 0.04 

I 0.04 


0.10 
0.13 
0.16 

0.15 

0.13 

0.13 


0.43 
0.54 
0.53 

0.49 

0.40 




0.80 
1.04 
1.00 

1.00 

0.90 

0.83 

0.69 

































The above results indicate that the moisture-excluding prop- 
erties of mixtures 6 and 7 are higher than those of 4 and 5. 
However, taking into consideration the behavior and the general 
appearance of the paint films and the fact that mixtures 4 and 
5 exclude moisture almost as well as linseed oil, I am inclined to 
recommend them as paint vehicles. 

SUMMARY AND CONCLUSIONS 

The drying properties of lumbang bato and lumbang banucalag 
oils are comparable with those of linseed oil. 

Lumbang bato oil is very similar to linseed oil in its properties 
as a paint vehicle, and like linseed has certain disadvantages for 
use in red lead paints. 

Lumbang banucalag oil cannot be used as a paint vehicle, espe- 
cially with red lead; it dries into a paste. This is also true 
with lumbang bato containing 75 and 90 per cent lumbang banu- 
calag (mixtures 6 and 7, Table V) . 

Lumbang banucalag containing between 50 and 75 per cent 
lumbang bato (mixtures 4 and 5, Table V) will make a good 
vehicle for red lead. 

This work was carried on at the suggestion and under the 
direction of Mr. George W. Heise. 



ILLUSTRATIONS 



Plate I. Modified Gardner-de Horvath apparatus used in testing paint 
films. 

TEXT FIGURE 

Fig. 1. Drying rate of linseed and lumbang oils. 

245 



Ai'.iilar. R. H.. Linseed ami Li mkani; Ou.s.l [PHIL. .Ioikn. Sri.. XII. A, No. 5. 




PLATE I. MODIFIED GARDNER-DE HORVATH APPARATUS. 



THE CRATER LAKE OF TAAL VOLCANO ' 

By George W. Heise 

(From the Laboratory of General, Inorganic, and Physical Chemistry, 
Bureau of Science, Manila) 

ONE PLATE AND ONE TEXT FIGURE 

Previous to 1911 Taal was an active volcano, which had erupt- 
ed frequently during historic times. The last and, as far as 
known, the greatest eruption occurred in January, 1911. An area 
of approximately 230 square kilometers was affected with de- 
vastating violence, (7) and over 1,300 people were killed. Since 
that time the volcano has been practically inactive. 

The following description of the volcano was published by 
Adams (l) about a year before the eruption: 

Taal volcano is situated on an island in Taal or Bombon Lake. The 
island, on which are found a number of extinct cinder cones and the active 
crater, has been built near the center of the lake by late volcanic 
activity. * * * The main crater which is situated near the center is 
usually referred to as Taal volcano. It is approximately circular in form. 
The southwestern border of the crater rim rises to an elevation of 320 
meters, which is the highest point on the island. The lowest points on 
the rim are about 130 to 150 meters in elevation. The lowest points on 
the floor of the crater are about on a level with the water of Taal 
Lake. * * * 

Two lakes lie within this crater. They are usually called the yellow 
lake and the green lake. During the rainy season there is a third tem- 
porary red lake. The yellow lake receives the natural drainage of the 
crater. It appears to be shallow and is hot, but does not boil. The green 
lake gives off steam from its surface and near its southern border boils 
violently as if over a vent. A circular crater is located to the south of the 
green lake. On its floor there are several boiling mud spots from which 
but little vapor rises. On the south border of the yellow lake there is a 
cone, called the red cone, because of the color of its crater. It is broken 
down on the south side and drains around its eastern base into the yellow 
lake. A vent from which steam issues with great force occurs on its 
northern outer base. The yellow lake now extends to this vent but formerly 
was separated from it by a narrow isthmus. There is a remnant of an 
older, large crater rim which forms a crescentic ridge rising southeast of 
the yellow lake and curving around to the south of the green lake, passing 
between the green lake and the crater with the mud spots. 

Plate I, fig. 1, is from a photograph of the crater that was 
taken before the 1911 eruption. 

1 Received for publication June 9, 1917. 

247 



248 The Philippine Journal of Science 1917 

A peculiar feature of Taal Volcano is the fact .that the main 
floor of the crater before the eruption was very nearly at sea 
level and that, owing to the ejection of much material during 
the eruption, the crater floor is now much lower. The volcano 
was little altered in outward appearance in the eruption of 1911, 
but the crater proper was greatly changed. According to the 
description by Pratt (7) — 

The absence of vegetation and the smooth drifted surface of the ash 
covering which is almost white in the sunlight, give the island an appear- 
ance of a vast snow heap. The crater rim is unbroken and save for minor 
fissures and cracks is intact. * * * 

The interior of the crater has been transformed. * * * The well- 
known Green Lake and Yellow Lake, which were small bodies of water, 
one of which (Yellow Lake) was quite shallow, referred to in descriptions 
of Taal since earliest historic times, are gone. In the position of the 
former Green Lake there is a new one, the water of which appears milky- 
white, due to suspended solid matter. The level of this lake was on 
February 17 approximately 70 meters below that of the sea. Green Lake 
had stood 5 meters above sea level. Two streams of hot water, the 
combined flow of which was estimated at 100 to 150 cubic meters per 
minute, were pouring into the lake. These streams came out of the crater 
walls about 50 meters above the lake level, seeping from just over a layer 
of fine-grained, impervious, bedded tuff. On the west shore of the lake 
a conical rock 50 to 70 meters in diameter rose to a height of 115 meters 
above the lake level. The upper 50 meters of this natural obelisk appeared 
to be bedded tuff, but the lower portion is massive basalt. A week later, 
the streams pouring into the crater lake had increased both in volume 
and in number, and the lake itself had risen apparently about 5 meters. 
* * * 

Although hot and heavily mineralized, the water which is flowing into 
the new crater probably is seepage from Lake Bombon through the crater 
walls. Since Lake Bombon stands a few meters above sea level, the new 
crater lake will probably rise in time to about sea level. 

A comparison of the old and new craters is shown in fig. 1. 

As predicted, the water entering the crater formed a single 
new lake, leaving no trace of the small lakes previously present 
(Plate I, fig. 2). This new lake, which is over a kilometer in 
width and at least 70 meters deep, has no visible inlet nor outlet. 
The following description was written by Gates, (6) in 1914: 

It [the crater] is about 2.3 kilometers long and 1.7 kilometers wide at the 
top. More than half of the bottom is occupied by a lake, whose elevation 
is about 2.5 meters above sea level, the same as that of the surrounding 
Lake Bombon. The water of the crater lake is clear, although dark colored, 
and salty. Its temperature decreased from about 37° C. in October, 1913, 
to about 32° in April, 1914. Swimming in it, although much like salt 
water bathing, was of course more exciting. Very little steam, if any, 
arose from the lake in either October or December, 1913, but in April, 
1914, some steam was noticed arising from a few places along the shore 
of the lake, as well as from small vents in the north crater wall, both 



Heise: Crater Lake of Taal 



249 




250 The Philippine Journal of Science 1917 

inside and outside the crater. From certain points on the crater rim 
sulphurous odors are noticeable, but none were detected in the bottom of 
the crater. Steep precipitous walls formed the boundary of the crater on 
all sides. At the foot of the walls, especially on the east side, large 
quantities of ash and mud have been washed down and have accumulated. 
The crater rim is highest on the south and north sides with altitudes of 
304 and 230 meters, respectively. Nearly all of the west side is low, the 
minimum elevation being about 95 meters. There are other low points 
on the east side. 

The above description holds very well for the conditions in 
the crater at the present time. There is a faint "sulphurous" 
smell at various points, due to mineral decomposition, and there 
are a number of small steam vents, principally on the north 
crater wall and at the west shore of the lake. Most of these 
vents were not readily accessible at the time the crater was 
visited, but on the northwest shore there were some very small 
openings, from which faint wisps of steam issued, and places 
where the sand, under water, was hot to the touch. As evidence 
of the insignificance of the present activity, it may be mentioned 
that the temperature of the lake was about 30° C, which is only 
a few degrees above that of ordinary surface waters in the 
Luzon lowlands. 

There are also a few steam vents at various points on the outer 
slope of the volcano, and there is at least one under water, on the 
northern shore of the island. These, too, are insignificant. The 
presence of steam vents under water has given rise to the belief 
that there are hot springs at various points. So far as could be 
determined, there is no foundation for this belief. 

Crystals of calcium sulphate, of iron salts, and of sulphur are 
found on the walls and floor of the crater, the first being abun- 
dant, the last very rare. Calcium sulphate is very plentiful, 
especially near the shore of the crater lake, where it has ap- 
parently been deposited from the water in large sheets a half 
centimeter in thickness. 

Analyses of the waters of the crater previous to the 1911 
eruption have been published by Centeno, in 1885,(4) and by 
Bacon, in 1906(2) and 1907.(3) The analysis of a sample of 
water from the stream flowing into the crater soon after the 
eruption was published by Cox. (5) 

Bacon (4) made qualitative tests for radioactivity in the 
waters and their sediments and found that the water and sedi- 
ment from the old "green lake" were very feebly active, the 
other waters and sediments showing no activity. Wright and 
Heise(8) were unable to detect the presence of radium in the 
water of the present crater lake. 



xii. a, 5 Heise: Crater Lake of Taal 251 

The results of the most recent analyses of samples taken by 
me in February and April, 1917, are given in Table I, coupled, 
for comparison, with the older analyses. The analysis of the 
water from Lake Bombon is also shown. 

A comparison of the above analyses shows that the crater 
waters became progressively more salty before the eruption ; that 
immediately after the eruption, as might be expected, the water 
entering the crater was much less highly mineralized ; and that 
the salt concentration has greatly increased since that time. 

The crater lake is at approximately the same level as Lake 
Bombon, that is, a few meters above sea level, and must be over 
70 meters deep. There are no indications of any springs, or of 
ingress of water other than that which would normally occur in 
an excavation extending below the water table. The crater is, 
in effect, a huge basin exposed to insolation. 

There is, however, a large volume of water added to the lake 
through the rainfall, which, in this region, averages about 1,900 
millimeters (75 inches) per year. The rainwater, running down 
the inner slope of the crater and into the lake, leaches the soluble 
salts from the volcanic ejecta forming the crater walls, increasing 
the amount of salts in the water; the intense tropical sunlight 
stimulates evaporation. Thus the lake water becomes more 
concentrated. This alternate addition of salt and evaporation 
of water is sufficient to account for the increasing concentration 
of the lake water. 

The quantity of soluble salts in the soil of the volcano may be 
inferred from the fact that large areas on the outer slopes are 
still practically unvegetated, except for occasional tufts of coarse 
grass, and that the inside of the crater is barren, except in 
isolated places. On the outer slope of the volcano, the soil, 
5 to 20 centimeters below the surface, still shows, six years 
after the eruption, large quantities of soluble matter. Analyses 
of the water-soluble constituents of two soil samples — one from 
a typical unvegetated area on the west slope of the volcano and 
one from a typical grass area from the west shore of the island — 
are given in Table II. For comparison, two analyses (5) of the 
water-soluble material in fine ash and ejecta, collected soon 
after the eruption, are shown. 



252 



The Philippine Journal of Science 



Table I. — Analyses of 

[Results expressed as gram? per 





Year. 


1885 


1885 


1906 


1906 




Centeno' 41 ... 
YellowLake* 


Centeno' 4 ' ... 
Green Lake". 


Bacon' 2 > 

Green Lake- 
Green 

1.1062 

1.128 N 

6.14b 

4.65b 

15.8678 

15.9312 

0.0634 


Bacon' 2 ' 

Yellow Lake. 

Yellow 

1.1763 

2.08 N 

7.57b 

25.7235. 

30.3175 

4.5940 




















0.16b ___ 


















2 - 6988 - 


6.0022 






0.0640. 




Iron and aluminium oxides (Fe203 + 
AI2O3). 












1.031 

0.7150 

0.3160 


1.711 

1.3250- 

0.386 




0.1315 


0.3036 

0.0612 






Aluminium (Ai) 






1.1892 

0.0934 

0.9432 

1.7217. 

0.0418 

6.3143 


3.485. 

0.1042 

1.8210 

0.9343 

0.0514 

11.2760 




0.0150 

0.0264 

0.6471. 


0.0136 

0.0335 

1.2405 








Chlorides (CI) 










Iodides (I) 












0.4122 

0.0396 


0.4186 

0.0515 


4.1303 

0.0273 


5.6768 

0.0391 




Loss on evaporating to total solids (?) 
































~ 







"Analytical results are given as recalculated by Bacon. ( : ) 

b Per cent. 

c Analyzed by J. Gonzalez Nunez, chemist, Bureau of Science. 

d Normal carbonate (as Na 2 C0 3 ), 24; bicarbonates (as HC03), 150. 



Heise: Crater Lake of Taal 



253 



waters from Taal Volcano. 
100 cubic centimeters except as noted.] 



Year. 


1907 


loo; 


1907 


1911 


1917 


1917 


Bacon' 3 '.. 

Boiling crater 
lake. 

Light grayish 
green. 


Bacon' 3 ' 

Green pool, 
north of boil- 
ing lake. 


Bacon' 3 ' 

Green Lake.. 

Green 






Lake Bom- 
bon (parts 
per mil- 
lion).d 

Nil. 

Nil. 

1540. 
75. o 
10. 

0.14. 

0.95. 

64. 

50. 

'580. 

194. 


Stream 


Crater Lake<=. 










" 

1.335N... 


1.78N 


0.0069N 


0.0147N 

0.0100N 






"" 






8.576 


















"" | 






0.0025= 

0.0410* 


! ;;;;;;;;;; 




0.07108— 






0.5157b 


0.9210 b 

0.8367 b 

0.0843b 


1.3446b 


0.01720 - 


0.0135 














0.00799 

0.00261 

0.05568 

0.09093 

0.25843 

0.02374 

0.60243 






0.8978 b 

0.2082 b 

0.4343 b 

0.7192b 

0.0048b 

4.8925b 


2.0927b 

0.1328 b 

0.1514b 

2.3246b_ 

0.0104 b 

10.9312 b 


0.0086 

0.0860 

0.2650 








0.0125b 




1.830.... 

Trace 

0.00005. 

0.330 














1.9688b 


2.3542 b 


0.27320 




2.1171b 


0.8284b 






0.6000s 






None 








Small 


do 



















' Determined turbidimetrically. 

f Determined by evaporation with hydrofluoric acid. 

* Loss on ignition. 



254 



The Philippine Journal of Science 



Table II. — Analyses of water-soluble constituents of Taal ejecta and soils. 

[Numbers give percentages by weight of air-dried sample.] 





1911 


1917 


Ash.' 4 ! 


Ejecta. «> 


Soil from 

unvege- 

tated 

area. 


Soil from 
grass 
area. 




0.82 
0.01 
0.16 

none 
1.61 
1.34 

trace 
0.30 

0.74 


0.95 
trace 

0.32 
none 

2.03 

1.26 
trace 

0.60 
none 

0.95 


0.12 
0.03 
3.81 


0.03 
nil 

0.04 


























4.33 


0.11 






0.03 
0.16 

8.75 


0.02 
0.04 
0.88 








i 



Apparently the two sets of analyses are not strictly com- 
parable because of the length of time between them and because 
of the probability that somewhat different materials were tested. 
It is significant, however, that the amounts of chlorides, that is, 
of the more soluble substances, are present in much smaller 
quantities in the 1917 samples than in those of 1911, indicating 
the extent to which leaching has taken place. Sodium and potas- 
sium, though not quantitatively determined in the later analyses, 
could not have been present as more than traces. The difference 
between the two 1917 analyses is of interest and furnishes at 
least a partial explanation for the differences in vegetation on 
different portions of the island. The acidity and the high soluble 
salt content, over six years after the eruption, indicate that a 
still greater salt concentration in the crater lake may be expected. 

REFERENCES 

1. Adams, G. I. Geologic reconnaissance of southwestern Luzon, Phil. 

Journ. Sci., Sec. A (1910) <; 5, 57-116. 

2. Bacon, R. F. The waters of the crater lakes of Taal Volcano, with a 

note on some phenomena of radioactivity, ibid. (1906), 1, 433-437. 

3. Idem. The crater lakes of Taal Volcano, ibid., Sec. A (1907), 2, 115-127. 

4. Centeno, J. Estudio geologico del volcan de Taal. Madrid (1885). 

5. Cox, A. J. The composition of the fine ejecta and a few other inorganic 

factors of Taal Volcano, Phil. Journ. Sci., Sec. A (1911), 6, 93-97. 

6. Gates, F. C. The pioneer vegetation of Taal Volcano, ibid., Sec. C 

(1914), 9, 391-434. 

7. Pratt, W. E. The eruption of Taal Volcano, ibid., Sec. A (1911), 6. 

63-86. 

8. Wright, J. R., and Heise, G. W. The radioactivity of Philippine waters, 

ibid., Sec. A (1917), 12, 145. 



ILLUSTRATIONS 



Plate I 

Fig. 1. Crater of Taal Volcano from the southeast, March, 1907. Photo- 
graph by Bacon. 
2. Panoramic view from the south rim of the crater of Taal Volcano. 
Photograph by Cortez. 

TEXT FIGURE 

Fig. 1. Superimposed cross sections of crater before and after the 1911 
eruption. 

255 



Husi:: Tin Cl 



Sri., XII, A. NO. 5. 




Heise: The Crater Lake op Taal Volcano.] 




[Phil. Joorn. Sa., \it. ,\. No. 5. 


8 ' ■■■MinnnMTi^iMBM i i^Mfc 


^■^^„ 




y ': - "1 




"^ 




*SKp IfijfilPK 


! 


~"~^^^l 


'■^s v,';^_ 


! 

1 


- ,--^Mi: 


IH^l . r^a^-- *-« j| 



Fig. 1. Crater of Taal Vclcano from the southeast, March, 1907. 




Fig. 2. Panoramic view from the south rim of the crater of Taal Volcano 
PLATE I. 



COMPOSITION OF BRICK AND MORTAR IN THE GREAT WALL 

OF CHINA • 

By J. C. Witt 

(From the Laboratory of General, Inorganic, and Physical Chemistry, 

Bureau of Science, Manila) 

ONE PLATE 

During a recent trip to northern China I visited the Great 
Wall at Shanhaikwan. At this point the wall is largely com- 
posed of gray brick laid with lime mortar. The bricks have a 
porous structure, somewhat resembling pumice, and are much 
larger than ordinary building bricks. They are so weak that 
pieces may be easily broken off with the fingers. The mortar 
is pure white, under the exposed surface, and is much stronger 
than the brick. 

The materials were of special interest to me because of recent 
research in ceramics and lime-burning at the Bureau of Science. 
A sample of each was taken and analyzed in this laboratory. 
The construction of the wall began in the third century before 
Christ, but was repaired eighteen centuries later. Therefore 
there was no means of knowing the age of the materials sampled, 
but apparently they were a part of the original structure. The 
general condition of the wall at this point is very good, as can 
be seen from Plate I, though near the top a number of the bricks 
are missing. Table I shows the analytical results. 
Table I. — Analysis of brick and mortar." 



Determination. 


Brick. 


Mortar. 




P.ct. 
0.10 
73.02 
18.96 


P.ct. 

43.88 
2.12 
0.44 

48.83 
4.03 
0.85 










1.05 
5.73 







3 Analyzed by F. D. Reyes, inorganic chemist, Bureau of Science. 

The brick is said to have been dried in the sun only. This 
was confirmed in the laboratory tests, because on ignition the 
material becomes dark red. If it had been originally burned 
in a kiln, the appearance of the wall would have been consid- 
erably different, and the strength and durability would doubt- 
less have been much greater. Both the general appearance and 
the analysis of the mortar indicate that no sand was mixed 
with the lime. It is apparent also that the stone from which 
the lime was made was of good quality. 



Received for publication May 16, 1917. 



257 



ILLUSTRATIONS 

Plate I 

Fig. 1. The Great Wall of China, showing location near top of wall at 
which samples were taken, January 4, 1917. 

2. View at base of wall. 

3. The wall extends over mountains and across valleys. 



Witt, J. C: Hru-k ash Mortak of GREAT \V. 



HI.. JOUBN. SCI., XII, A, No. 5. 






■ 


■ 


PM ^S 


■ 




jgj P^^^^^ 




| 




IB 


^^^^^^W 


!&i- 


■-■- 


j^&g&te^ 


^fSiil 


mjfc- 


■fr " r 


■ ' ■ .-jti.-.:;- ■ 



Fig. 2. View at base of wal 




Fig. 3. The wall extends over mountains and across valleys. 
PLATE I. 



SOME LIMITATIONS OF THE KJELDAHL METHOD » 
By Harvey C. Brill and Francisco Agcaoili 

(From the Laboratory of Organic Chemistry, Bureau of Science, 
Manila) 

It is an interesting fact that the various chemistry textbooks 
dealing with analytical methods 2 in describing the Kjeldahl 
method for the determination of the nitrogen of nitrogenous 
organic compounds cite only hyrazine, nitro, and similar nitrogen 
and nitro-oxygen groups as the exceptions, which require special 
treatment before their total nitrogen can be obtained by this 
process. 

Recent investigations have shown that compounds other than 
those enumerated in the above references are resistant to decom- 
position when subjected to treatment according to the Kjeldahl 
method or some of its modifications. Pyridine sulphonic acids 
are prepared by heating pyridine with sulphuric acid. 3 The yield 
is increased. by the addition of certain metallic sulphates. This 
increase results from the raising of the temperature of the 
sulphuric acid solution due to the presence of the sulphate and 
possibly from the catalytic action of the latter. But more of 
the sulphonic acid derivative is synthesized in the presence of 
the sulphate at the higher temperature than in its absence. 

1 Received for publication March, 1917. 

J In Bull. U. S. Dept. Agr. (1912), 107, 5, is the following in regard to 
applicability of the Kjeldahl and the Gunning methods: "Not applicable 
in the presence of nitrates." Sudborough, J. J., and James, T. C, in 
Practical Organic Chemistry, Van Nostrand Company, New York (1908), 
61, write: "It should be remembered the following groups of compounds 
do not yield quantitative results unless subjected to a preliminary treat- 
ment: Nitro, nitroso azo, hydrazo, diazonium compounds and probably 
cyanogen derivatives and platinichlorides of bases." Leach, A. E., Food 
Inspection and Analysis, John Wiley & Sons, New York (1914), 69, in 
discussing the use of the Kjeldahl and the Gunning methods, warns the 
analyst that: "Neither method in its simplest form is applicable in the 
presence of nitrates; if these are present, a modification must be used. 
The Gunning Arnold method is employed for the determination of the 
nitrogen in pepper, as the piperin is not completely decomposed by the usual 
processes." 

'Wiedell and Wurman, Monatsh. f. Chem. (1895), 16, 749. Meyer and 
Ritter, ibid. (1914), 35, 765. 

261 



262 



The Philippine Journal of Science 



Funk, 4 in his study of the vitamine bodies in rice polishings, 
found that he obtained smaller yields of nitrogen from these 
compounds when the Kjeldahl method was used than when the 
Dumas or absolute method was used. 

Table I. — Nitrogen determination by means of the Kjeldahl method. 



Compound. 


Nitrogen. 


Nitrogen 
liberated. 


Theory. 


Found. 




Per cent. 
11.96 

11.76 
20.90 

17.72 
16.47 

10.85 

10.85 

9.53 
32.54 

10.00 

17.34 

22.23 
35.44 


Per cent. 

11.92 

11.99 

11.75 

11. 79 

f 11. 62 

1 11.69 

17. 20 

16. 61 

12.18 

12.19 

12.34 

11.30 

9.50 

12.70 

8.41 

8.79 

8.58 

8.48 

6.28 

6.14 

32.44 

| 10. 12 

1 9.90 

r 16. 93 


Per cent. 
99.73 

100.00 
98.25 
98.69 
98.80 
99.40 
82.30 
79.47 
68.72 
68.73 
69.64 
68.63 
57.69 
77.12 
77.50 
81.00 
79.07 
76.37 
65.90 
64.43 
99.66 

100.00 
99.00 
97.67 


















Oxynicotinic acid 




[ 15.89 i 91.63 
f 22.25 100.00 




| 34. 95 

1 34. 99 

33.39 

28.02 


98.60 

100.00 
98.82 




33.34 
28.35 







Williams, 5 who was investigating this subject of compounds 
with antineuritic properties, at once saw the possibility of making 
use of this property of resisting decomposition to determine the 
chemical character of these compounds. As a preliminary step 
in the solution of this problem we undertook a determination of 



4 Funk, C, Journ. Physiol. (1913), 46, 177. Drummond, J. C, and Funk, 
C, Biochem. Journ. (1914), 8, 598. 

"Williams, R. R., This Journal, Sec. A (1916), 11, 49. 



Brill and Agcaoili: Kjeldahl Method 



263 



the nitrogen content of various classes of nitrocarbon com- 
pounds by means of the Kjeldahl method to determine what type 
of compounds yields only a part of its nitrogen by this process. 

A sample of approximately 0.2 gram of the pure compound 
was heated with concentrated sulphuric acid (20 to 30 cubic centi- 
meters), potassium sulphate (5 grams), and copper sulphate 
(0.5 gram) for two and one-half to four hours or for one-half 
hour after the solution had become clear, at the boiling point 
of the solution. The results of this examination are given in 
Table I. 

Benzylcyanide and phenylcyanate were included in this inves- 
tigation, as it was at first thought that the low yields of nitrogen 
obtained might be due to the breaking down of the molecule with 
the liberation of hydrocyanic acid. However, it seems more 
plausible to attribute the low yields of nitrogen from pyrrole, 
pyridine, piperidine, quinoline, isoquinoline, and oxyquinoline to 
the formation of sulphonic acid derivatives and their subsequent 
resistance to decomposition by the sulphuric acid. 

The use of vanadium oxide • has been recommended as a cataly- 
zer in the Kjeldahl method. The method was employed by us 
in the determination of the nitrogen of piperidine without satis- 
factory results. 

Table II. — Nitrogen determination of piperidine by use of the modified 
Kjeldahl method.* 



No. of analysis. 


Compound. 


Nitrogen. 


Total ni- 
trogen 
liberated. 


Theory. 


Found. 






Per cent. 
16.47 
16.47 
16.47 


Per cent. 
12.23 
12.15 
12.96 


Per cent. 
74.26 
73.77 
78.69 




do 




do 







* No 1 was prepared in the manner described by Wunder and Lascar ; No. 2 had 0.5 
gram of bismuth oxide substituted for the vanadium oxide ; while No. 3 had the same amount 
cf antimony oxide used as a substitute. 

The results obtained in the experiment using vanadium oxide 
are no better than where bismuth or antimony oxide was used. 

Dakin and Dudley 7 state that piperidine by prolonged heating 
with sulphuric acid in the presence of potassium sulphate and 
copper sulphate gives up all its nitrogen, but that pyridine gives 
less satisfactory results. In the determination of the nitrogen 



"Wunder, W., and Lascar, O., Journ. Phurm. et Chim. (1914), 19, 329. 
T Dakin, H. D., and Dudley, H. W., Journ. Biol. Chem. (1914), 17, 275. 



264 



The Philippine Journal of Science 



content of coal, Fieldner and Taylor 8 found that low results were 
always obtained when the digestion was stopped as soon as the 
solution became colorless. 

We have used piperidine with potassium sulphate and different 
amounts of mercuric oxide and with sodium sulphate and dif- 
ferent amounts of mercuric oxide without quantitative yields, 
even when the mixture was heated for a period of ten hours. 



Table III. — Nitrogen determination of pi-peridine 
method. 



modified Kjeldahl 



No. of sample. 


Sodium 
sulphate. 


Potas- 
sium 
sulphate. 


Mercuric 
oxide. 


Time 

heated. 


Nitrogen. 
Theory, j Found. 


Total ni- 
trogen 
liberated. 




Grams. 


8 


Grams. 
8 

8 



Grams. 
0.3 


Hours. 

10 


Per cent. 
16.47 
16.47 


Per cent. 
12.63 


Per cent. 
76.69 




0. 7 10 
0.3 8 























The above methods were unsuccessful in giving quantitative 
yields. 

Finally the method known as the Gunning-Arnold method, 
which is recommended by Leach, ° for the determination of the 
nitrogen in black pepper, was employed. This method gives 
good results with pyridine when the heating is continued for a 
considerable period after the solution becomes clear. An attempt 
was made to substitute sodium sulphate for the potassium sul- 
phate in the Gunning-Arnold method, but the substitution was 
unsuccessful. 

Table IV. — Nitrogen determination of pyridine by the Gunning -Arnold 
method and by a modified Gunning-Arnold method. 



um sul- | 



Grams. Hours. 



Theory. 


Found. 


Per cent. 


Per cent. 


17.72 


12.7 ' 


17.72 


12.0 


17.72 


13.6 


17.72 


15.8| 


17.72 


17.2 


17.72 


17.6 ! 


17. 72 


16. 6 1 


17.72 


16.3 ■ 



Total 

nitrogen 

liberated. 



76.2 

96.5 
99 3 
03. 6 
92.0 



•Fieldner, A. C, and Taylor, C. A., Journ. Ind. & Eng. Chem. (1915), 
7, 106. 
* Loc. ciL 



xii. a, 6 Brill and A gcaoili: Kjeldahl Method 265 

The results when sodium sulphate was used are in every case 
low. Digesting the pyridine for a longer time is conducive to 
more nearly theoretical yields. Where mercury is used, care 
must be taken to precipitate it as the sulphide. A loss in nitrogen 
is liable to arise from this source, because of the difficulty of 
decomposing the ammonio-mercurial compound. 10 

SUMMARY 

The Kjeldahl method gives low results for nitrogen with pyri- 
dine, piperidine, quinoline, isoquinoline, oxyquinoline, pyrrole, 
and in some cases with nicotine. The authors believe this arises 
from the formation of sulphonic acid derivatives and their resist- 
ance to decomposition. ' 

The Gunning-Arnold method gives more reliable results with 
pyridine when heated for a considerable period after the solution 
has become clear. 

Sodium sulphate is not conducive to good yields and cannot 
be substituted for potassium sulphate. 

10 Justin-Mueller, Ed., Bull. Sci. Pharm., Paris (1916), 23, 167. Nolte, 
Otto, Zcitschr. f. anal. Chem. (1916), 55, 185. 



PUBLICATIONS FOR SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS— Continued 



BOTANY 

A FLORA OF MANILA 
By Elmer D. Merrill 

Order No. 419. Paper, 490 pages, $2.50, 
postpaid. 
Practically a complete flora of the cul- 
tivated areas in the Philippines. Descrip- 
tions, with keys, of over 1,000 speoies, 590 
penera, and 136 families, with native names, 
glossary of technical terms, etc 



PHILIPPINE DIPTEROCARP FORESTS 
By William H. Brown and Donald M. 

MATTHBW3 

Order No. 432. Paper, 150 pages, 1 map, 
13 plates, and 12 diagrams, $1.00, 
postpaid. 
In Philippine Dlpterooarp Forests the 
authors present a very comprehensive discus- 
sion of the growth and development of dip- 
terocarp trees and of the other element* of 
lowland Philippine forests. 



INDO-MALAYAN WOODS 
By Fred W. Foxworthy 

Order No. 411. Paper, 182 pages, 9 

plates, $0.50, postpaid. 

In Indo-Maiayan Woods, Doctor Fox- 
worthy has brought together a large amount 
of accurate Information concerning trees 
yielding woods of economic value. 



ZOOLOGY 

A LIST OF MAMMALS OF THE 

PHILIPPINE I8LANDS, EXCLU- 

SIVE OF THE CETACEA 

By Ned Hollister 

Order No. 418. Paper, 64 pages, $0.50, 

postpaid. 

The distribution of each speoies Is given, 
and the original descriptions are oited. 



Z OOLOGY— Continue <I 

A MANUAL OF PHILIPPINE BIRDS 

By Richard C. McGregor 

Order No. 103- Paper. 2 parts, 769 

pages, $4, postpaid. 

A Manual of Philippine Birds contains 
In oompact form descriptions of all the 
known speoies of Philippine birds. The usual 
keys and diagnoses of orders, families, and 
genera help the novioe In identification. 



A CHECK-LIST OF PHILIPPINE 



By David Stabb Jordan and Robert Eabi. 
Richardson 



Paper, 7S pages, $0.75, 



This list will be found a convenient guide 
to the synonymy of Philippine ichthyology. 
The nomenclature Is thoroughly revised, and) 
the distribution of eaoh speoies within the 
Philippine Islands is given. 



A CATALOGUE OF PHILIPPINE 
COLEOPTEBA 



By W. Schultzb 



198 pagee, $1.00, 



This catalogue Includes the names of all 
species of Coleoptera that have been reoorded 
from a definite locality In the Philippine 
Islands. References to original description* 
and other Important notes are given. The 
economio appendix lnolude6 oomment on 
those species of beetles which are known to 
be injurious or beneficial to man. 



PRICES ARE IN UNITED STATES CURRENCY 

Orders for these publications may be sent to the BUSINESS MANAGES, 
PHILIPPINE JOURNAL OF SCIENCE, BUREAU OE SCIENCE, MANILA, P. L, 
or to any of the agents listed below. Please give order number. 



The Macmillan Company^ 1 64—66 Fifth Avenue, New York, TJ. S. A. 

Wm. Wesley & Son, 2 8 Essex Street, Strand, London, W. C, England. 

Martlnus Nijhoff, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdinandstrasse 2, Berlin N. W., Germany. 

Kelly & Walsh, Ltd., 32 Raffles Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 19 Baillie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



CONTENTS 

)WN, WILLIAM H., and ARGuELLES, ANGEL w 
composition and moisture content of the soils in the types of 
vegetation at different elevations on Mount Maquiling 221 

AGUILAR, R. H. A comparison of linseed oil and lumbang oils 

as paint vehicles 235 

HEISE, GEORGE W. The crater lake of Taal Volcan: 247 

WITT, J. C. Composition of brick and mortar in the Great Wall 
of China ... 

BRILL, HARVEY C., and AGCAOILI, FRANCISCO. Some limi- 
tations of the Kjeldahl method 261 



u. s. 

currency. 

$2.00 

S.OO 

2.00 

2.00 

5.00 

7.00 

.50 



The "Philippine Journal of Science" is issued as follows: 

Section A. Chemical and Geological Sciences and the Industries.. 

Section B. Tropical Medicine 

Section C. Botany 

Section J>. General Biology, Ethnology, and Anthropology (Sec- 
tion D l>egan with Volume V) 

Entire Journal, Volume II, III, IV, or V 

Entire Journal, beginning with Volume VI ... 

Single numbers (except of Volume I) - 

Each section is separately paged and indexed. 
Authors receive 100 copies of their papers free. 

Volume I, 1906 (not divided into sections) and supplement, sold 
only with a complete file of section A, B, or C _. 

Supplement to Volume I (botany) 

Volnme I (without supplement), sold only with a complete flic of 
section A, B, or C 

Single numbers of Volume I - 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, P. I., or to any of the agents 
listed below: 

AGENTS 

The Macmillan Company, 64—66 Fifth Avenue, New York City, U. S. A. 

"Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martinns Nijhoff, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prlnz Louis Ferdlnandstrasse 2, Berlin, N. W., Germany. 

Kelly & Walsh, Limited, 32 Baffles Place, -.raits Scttlen. 

A. M. & J. Ferguson, 19 Baillie Street, Colombo, Ceylon. 

Thacker, Spink & Co., P. O. Box 54, Calcutta, India. 



[0.00 
S.50 



e.5o 

.75 



second-class matter. 



\II, Sec. A, No. 6 



November, 1917 



THE PHILIPPINE 

JOURNAL OF SCIENCE 



ALVIN J. COX, M. A., Ph. D. 

GENERAL EDITOR 



Section A 

CHEMICAL AND GEOLOGICAL SCIENCES 
AND THE INDUSTRIES 



EDITED WITH THE COOPERATION OF 

H. C. BRILL, Ph. D.; J. R. WRIGHT, Ph. D.; G. W. HEISE, M. S. 

J. C. WITT, Ph. D.; T. DAR JUAN, A. B.; A. H. WELLS, A. B. 

R. C. MCGREGOR, A. B.; H. E. KUPFER, A. B. 




MANILA 
BUREAU OP PRINTING 

1917 



PUBLICATIONS FOE SALE BY THE BUREAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS 



ETHNOLOGY 

A VOCABTTLABY OF THE IGOBOT LAN- 
GUAGE AS SPOKEN BY THE 
BONTOC IGOBOTE 

By Walter Clayton Clajt 

Order No. 40S. Paper, 89 pages. $0.75. 

postpaid. 
The vocabulary is given in Igorot-Englioh 
and English-lgorot. 



THE NABALOI DIALECT 

By Otto Schebkix 

and 

THE BATAKS OE PALAWAN 

By Edward Y. Miller 

Order No. 403. Paper, $0.25; half rno- 

rooco, $0.75; postpaid. 

The Nabaloi Dialect (65 pages, 29 

plates) and the Bataks of Palawan (7 

pages, 6 plate*) are bound under one cover. 



THE BATAN DIALECT AS A KEHBEB 

OF THE PHILIPPINE GBOTJF 

OF LANGUAGES 

By Otto Scheerer 

■'-■; ''■ and 



By Carlos Everett Conant 

Order No. 407. 

These two papers are issued under one 
cover, 141 pages, paper, $0.80, postpaid. 



ETHNOLOGY— Continued 



By Najbeb M. Saleujv 

Order No. 405. Paper, 107 pages, 16 

plates, 5 diagrams, $0.25; half mo- 
rocco, $0.75; postpaid. 
TMIs volume deals with the earliest 
written records of the Moros in Mindanao. 
The names of the rulers of Maoin 
recorded in five folding diagrams. 



NEGBITOS OF ZAOKBALES 

By William Allan Beep 

Order No. 402. Paper, 83 pages, 62 

plates, $0.25; half morooco, $0.75; 
postpaid. 
Plates from photographs, many of whioh 
were taken for this publication, show orna- 
ments, houses, men making fire with bamboo, 
bows and arrows, danoes, and various type* 
of the people themselves. 



INDUSTRIES 

PHILIPPINE HATS 
By C. B. Robinson 

Order No. 415. Paper, 66 pages, 8 

plates, $0-50 postpaid. 
This paper Is a concise record of the 
history and present condition of hat making 
In the Philippine Islands. 



THE STTBANUNS OF 8INDANGAN BAT 

By Emerson B. Christie 

Order No. 410. Paper, 121 pages, 1 

map, 29 plates, $1.25, postpaid. 

Sindangan Bay Is situated on the north- 
ern ooast of Zamboanga Peninsula. The Su- 
banuns of this region were studied by Mr. 
Christie during two periods of five and six 
weeks, respectively. 

The 29 plates illustrate the Subanuns at 
work and at play; their industries, houses, 
altars, and implements; and the people 
themselves. 



By Herbert S. Walker 

Order No. 412. Paper, 145 pages, 10 
plates, 1 map, $1.25, postpaid. 

Considered from the viewpoint of prac- 
tical utility, Mr. Walker's Sugar Industry 
In the Island of Negros Is one of the most 
Important papers published by the Bureau 
of Scienoe. This volume is a real contribu- 
tion to the subject; it Is not • mere com- 
pilation, for the author was in the field and 
understands the oondltions of which he 
writes. 



THE HISTOBY OF BTTLTT 

By Najreb M. Saleebt 

Order No. 406. Paper, 275 pages, 4 
maps, 2 diagrams, $0.75, postpaid. 
In the preparation of his manuscript for 
Tho History of Sulu, Doctor Saleeby spent 
much time and effort in gaining access 
to documents In the possession of the Sultan 
of Sulu. This book is a history of the 
Moros In the Philippines from the earliest 
timet to the American oocu atioi.. 



By Charles S. Banks 

Order No. 413. Paper, 53 pages, 20 
plates, $0.75, postpaid. 
In A Manual of Philippine Silk Culture 
are presented the results of several years' 
actual werk with silk-producing larva to- 
gether with a description of the new Philip- 
pine raoe. 



THE PHILIPPINE 

Journal of Science 

A. Chemical and Geological Sciences 
and the Industries 

VOL. XII NOVEMBER, 1917 No. 6 



ALCOHOL FROM DISCARD MOLASSES IN THE PHILIPPINE 
ISLANDS l 

By Harvey C. Brill and Leavitt W. Thurlow 
(From the Laboratory of Organic Chemistry, Bureau of Science, Manila) 

The utilization of the discard molasses of the sugar mill is 
yearly becoming of more importance. Several uses suggest 
themselves. 

1. Molasses has value as a cattle food. Its application to this 
use depends on local conditions of freight rates and whether it 
is to be transported long distances or consumed near where it 
is produced. At present very little is thus utilized in the Philip- 
pine Islands, although this use may have possibilities, for in the 
United States molasses for this purpose, inferior to the Philip- 
pine products, sells at a greater price even though it is shipped 
considerable distances. 

2. Attempts have been and are being made to use it for fuel. 
In Honolulu it has been burned to recover the ash for fertilizer 
purposes. All the phosphates and potash, except a small vola- 
tilized portion, remain. Practically all the nitrogen is lost when 
molasses is burned. It is difficult to handle as a fuel, because of 
easily fusible constituents of the ash that form a glaze on the 
walls of the furnace. However, since fuel is scarce in many 
centrals, the use of extra grate bars, effort, and expense are 
justifiable in the utilization of molasses as fuel. 

3. Its use as a binder in paving brick is still in the experi- 
mental stage, and its usefulness for this purpose cannot be 
predicted at this time. 

4. It has been used as a binder in the manufacture of bri- 
quettes from coal dust. Only a portion of it could be utilized 

1 Received for publication May, 1917. 
151034 267 



I 



268 The Philippine Journal of Science mi 

in the manufacture of briquettes were the enterprise ever so 
successful, and as coal dust 2 promises to be extensively used per 
se for fuel purposes, no great quantities of molasses are likely 
to be used for the manufacture of briquettes from coal dust. 

5. Molasses is valuable per se as a fertilizer, but in this use 
too much sugar is wasted and the effects of the sugar on the 
soil are not altogether beneficial. The bacteria s that fix nitrogen 
in the soil independently of the host plant are stimulated by su- 
gar, especially glucose. One should expect to be able so to 
stimulate the bacteria by the u^e of molasses that the nitrogen 
fixation from this source would become important. All attempts 
to do this have been disappointing, since the sugar likewise 
stimulates the class of bacteria that breaks down the stored-up 
nitrogen compounds. 

6. If the molasses were first fermented, it would yield a profit 
from the alcohol obtained and the lees would become a valuable 
source of fertilizer, because of their nitrogen, potash, and phos- 
phate content. The making of alcohol from molasses and the 
recovery of the fertilizing ingredients is the most profitable of 
the known uses for molasses. 

All the fertilizing ingredients ordinarily removed from the 
soil by the cane are concentrated in the molasses, so that when 
these are recovered and returned to the soil, together with the 
ash from the fiber or bagasse, the soil has suffered no loss and 
theoretically is as capable of producing a second crop as it was 
of producing the preceding one. The sugar industry is unique 
in that the sugar produced represents no constituent taken 
from the soil that must be returned in the form of fertilizer. 
Consequently, if the mineral ingredients found in the bagasse 
and in the molasses are returned to the soil, the soil is no more 
depleted than before the crop was harvested. Discard molasses 
in the Philippine Islands sells for about 6 centavos 4 a gallon 
(1.6 centaVos a liter), or about 10 pesos a ton. Peck and Noel 
Deerr 5 have estimated the value of the fertilizer ingredients in 
1 ton of Hawaiian molasses at 14 pesos and the cost of the re- 
covery of these constituents at 6 pesos. 

* The Chicago & North Western Railway Co. is successfully using coal 
dust in running one of its engines. It claims to obtain higher efficiency 
from coal dust than from lump coal. 

3 These are a different group from those existing in the nodules of 
legumes. 

* One peso Philippine currency equals 100 centavos, equals 50 cents 
United States currency. 

'Bull. Hawaiian Exp. Station, 26, through Sugar (1914), 16, No. 7, 38. 



xii. ,\. c Ihlll and Tkurlow: Alcohol from Molasses 269 

Table I. — Fertilizer constituents of Philippine and of Hawaiian molasses. 



Constituent. 


Philip ' Hawai- 
pine. i ian. 




Per cent. 

1.39 
0.38 

0.21 


. 3.S9 
0.21 
0.G4 









The potash and nitrogen content are greater for Hawaiian 
. molasses than for Philippine molasses. The Hawaiian cane 
takes up more of these constituents, because the quantity avail- 
able is greater due to the general use of fertilizers in the 
Hawaiian Islands. In many cases no fertilizer is added to the 
sugar lands of the Philippine Islands and no attempt is made 
to return the fertilizer ingredients found in the ash of the 
bagasse and in the molasses. Thus it is only a question of years 
before fertilization must be practiced, or this land will become 
exhausted and the planters' loss will be much larger than at 
present on account of the smaller crops resulting. Therefore 
it is expected that the fertilizer value will never be lower than 
now. 

In practice about 1 ton of molasses is produced for every 5 to 6 
tons of centrifugal sugar. During 1915, 211,012,817 kilograms 
of sugar were exported from the Philippine Islands. Were 
this exportation all centrifugal sugar, it would represent over 
30,000,000 kilograms of molasses. If this molasses were con- 
verted into alcohol and the fertilizer ingredients were re- 
covered from the lees, they would be worth more than 125,000 
pesos at the price quoted for the fertilizer ingredients in 
Hawaiian molasses," while the alcohol would represent a value 
of, approximately, 1,750,000 pesos calculated at 17.2 centavos, the 
current (December 1, 1916) selling price per liter of 182 proof 
denatured alcohol. 

A considerable portion of the Philippine discard molasses is 
now being used for the manufacture of alcohol, and several con- 
cerns are planning to extend their activities so that the waste 
here is not greater than in many other sugar-growing countries. 
The methods of fermentation are crude and capable of much im- 
provement. For example, no well-directed effort is made to 
keep an accurate control of the percentage yields by deter- 
mining the sugar content of the molasses and by diluting the 
wort to a definite strength. Care is seldom taken to keep the 
vats and machinery clean or to sterilize the vats, and 'even com- 



Peck and Deerr, loc. cit. 



270 The Philippine Journal of Science im 

mon cleanliness is often strikingly lacking. Neither the inocula- 
tion of the wort with a culture of yeast in order that the yeast 
may have a start on the bacteria, nor the distillation of the fer- 
ment when the alcohol content is at a maximum, is always 
carried out. It is common practice to allow the wort to become 
inoculated with yeast from the air. From one to three days are 
required to bring about active fermentation, and as a consequence 
the ferment becomes infested with bacteria, which destroy much 
of the sugar before the yeast crowds them out. No attempt 
is made to use good water for diluting the molasses, and it is a 
common practice to use dirty water, thus introducing large 
quantities of bacteria. Poor yields are obtained when poor 
methods are used. When the vats and solutions are sterilized, 
when pure cultures are used, and more care is taken, the cost of 
handling the solutions will increase, but the greater yield will 
more than compensate for this. Some figures presented by 
Owen 7 on the relation of profit to efficiency of fermentation are 
significant enough to be quoted here. 

Table II. — Profits from the fermentation of molasses, under varying 

conditions of efficiency. 

Theoretical yield. Profit per gallon. 

Per cent. Cents." 

57 -5.5 

60 0.1 

66 1.43 

69 2.0 

75 3.3 

80 4.3 

85 5.4 

90 6.5 

94.7 7.4 

" United States currency. 

This estimate is based on a cost price for Cuban molasses 
(total sugar as glucose, 55.11 per cent) of 10 centavos a gallon 
and a selling price of alcohol of 76 centavos a gallon, 180 proof. 
Theoretically 51.1 per cent of the weight of the sugar is the 
weight of the resulting alcohol, but in practice this cannot be 
attained, since some of the sugar is consumed by the yeast and 
some of it is converted into glycerol, succinic acid, cellulose, etc., 
so that the highest percentage possible is 94.7 per cent of the 
theoretical. 

Table II shows that no profit results when a yield below 60 
per cent is obtained from molasses under these conditions of cost 
and selling price. A profit of 2 centavos is not obtained until 
we reach an efficiency of 65 per cent. At 80 per cent, which is 

'Sugar (1914), 16, No. 7, 32. 



Brill and Thurloiv: Alcohol from Molasses 



271 



considered good distillery practice, the profits increase to 8.6 
centavos ; while at 94.7 per cent of the theoretical, which is the 
highest possible yield obtainable, the profits have risen to 14.8 
centavos a gallon, almost three times as great as the profits 
for a 75 per cent yield. This seemingly high increase in the 
profits is due to the increased yields more than compensating 
for the increased cost of production. In the Philippines, where 
the molasses 8 costs 6 centavos a gallon and 182 proof alcohol 
sells at 17.2 centavos a liter, the profits resulting are as given 
in Table III. 

Table III. — Profits from the fermentation of Philippine molasses under 
varying conditions of efficiency. 







Theoretical 
yield. 


Profit per gallon of 
molasses. 


Calculated 

on the basis 

of Table 

II. 


Based on ! 
the estima- 
ted cost of 
alcohol pro- 
duction in 
the P. I. » 


Per cent. 


Centavos. 


Centavos. 


57 


5.8 


4.5 


60 


7.0 


5.7 


65 


8.9 


7.6 


70 


10.8 


9.5 


75 


12.7 


11.4 


80 


14.6 


18.3 


85 


16.6 


15.3 


90 


18.5 


17.2 


94.7 


20.3 


19.0 



» Calculated on the basis of a cost of 17.3 centavos for converting 1 gallon of molasses into 
alcohol. 

The cost of converting 1 gallon of molasses in the United 
States into 90 per cent alcohol is 16 centavos. The estimate as 
deduced from certain data collected in the Philippine Islands 
makes the cost 17.3 centavos, only slightly greater than the 
above, and leaves a good margin of profit for the manufacturer. 
The larger profit for alcohol produced in the Philippine Islands 
is due to the smaller price paid for the molasses and its higher 
sugar content. A yield of 80 per cent of the theoretical gives 
a profit on the molasses of 13.3 centavos per gallon, while a 60 
per cent yield allows a profit less than half as great. Any im- 
provement in the methods of fermentation and distillation of 
molasses would mean an increase of revenue to the Islands and 
larger profits to the manufacturers. No alarm need be felt in 

8 Average total sugar as glucose for 10 samples, 62.6 per cent. 



272 The Philippine Journal of Science 1917 

regard to a possible overproduction of alcohol ° in the future. 
It is certain to have an increased sale for power purposes with 
the rise in the price of gasoline. Special carburetors are sure 
to be invented which will be adapted for alcohol or gasoline 
interchangeably as a fuel. The data in Table IV are suggestive 
of what will take place on a large scale at no distant date. 

Table IV. — Data obtained in Germany with various fuels and mixtures on 
a model 191 U Mercedes car, ordinary carburetor." 



1 Distance 
F(] _i i Speed, 1 covered 
* uei - per hour, on 1 pint 

, of fuel. 


Parts. 


Miles. I MiUs. 
42 4.66 








39 4. 34 




38 

42 
44 


4.10 
3.72 
3.79 
3.60 











• Chem. Eng. & Manufact. (1916), 24, 86. 

At the present price 10 of these fuels, mixtures of alcohol 
with benzol are economical sources of power. 

England " produces 100 proof alcohol at a cost of 19.5 cen- 
tavos per liter; Germany produces absolute alcohol at a cost of 
from 12.5 to 15 centavos per liter; the United States at 16 
centavos a liter of 182 proof ; and the Philippine Islands at 17.3 
centavos a liter of 182 proof. 12 To compete with alcohol pro- 
duced by Germany or the United States, the Philippine Islands 
must establish efficient methods in order to reduce the cost of 
production. 

Improvements in the process of fermenting molasses must 
proceed along the lines of the concentration of the ferment, 
control of the temperature during fermentation, the kind of 
yeast used, including freedom from bacteria and wild yeast, and 
the choice and use of yeast foods and stimulants. 

As fermentation is the result of the activity of living organ- 
isms, a study of the most favorable conditions for their growth 
and activity should result in the discovery of the environment 

"See Cox, Alvin J., This Journal, Sec. A (1909), 4. 232. 

10 Cost of 90 per cent benzol per liter, 31.7 centavos; gasoline, 33 cen- 
tavos; alcohol, 182 proof, 17.2 centavos. (December 1, 191C.) 

11 Martin, Industrial Chemistry, Organic. D. Appleton and Company, 
New York (1913), 280. 

a Calculated from data furnished by local distillers. 



xji.a. 6 Brill and Thurlow: Alcohol from Molasses 273 

most suitable for them and should result in an accompanying 
benefit to the distiller in the increased yields of alcohol. 

In our search for a suitable yeast we were assisted by Mr. 
K. B. Graae, bacteriologist, Bureau of Science, who isolated 
five cultures of yeast from fermenting nipa juice. These after- 
ward proved to belong to the same race of top-fermenting yeast, 
a Saccharomyces. The cells are smaller than those of race II 
or race XII 13 and are fairly quick-fermenting. The use of a 
quick-fermenting yeast is an advantage, since owing to its rapid 
growth it crowds out bacteria and wild yeasts. 

For the growth of yeast, properly adjusted foodstuffs are 
necessary. A solution of pure cane sugar would undergo 
fermentation, but the yeast would not grow and the fer- 
mentation would stop after the enzymes already formulated by 
the yeast became exhausted. Yeasts require for their growth 
certain organic and inorganic compounds, such as nitrogen 
bodies (ammonium salts, amides, peptones, phosphates, sulphates, 
potassium, and magnesium salts), which are always present in 
available form in sufficient quantities in grape juice and beer 
wort, but not in molasses. The presence of proper food 
stimulates yeast growth, and rapid growth of cultured yeast 
prevents the development of bacteria and wild yeast. The 
latter are undesirable, since some of them convert the sugar 
into substances other than alcohol and thus lessen the yield, 
while others produce alcohol, but in reduced amounts. Yeast 
can be accustomed to the presence of the' antiseptics, so that 
they withstand them when present in not too large quantities. 14 
Advantage is taken of this fact to keep the ferment free from 
bacteria and wild yeast, which are much more susceptible to 
the influence of such reagents.^ Sulphuric acid and salts of 
hydrofluoric acid are extensively used for this purpose. 

A 5 per cent solution of alcohol is strong enough to stop the 
propagation of many yeasts, but the enzymes continue to act for 
some time after the propagation of the yeast has stopped. 
Others by cultivation have become accustomed to more concen- 
trated solutions ; consequently they will ferment in much stronger 
solutions. A concentrated solution of molasses ferments more 
slowly after a certain period, because of this higher percentage 

13 Race II and race XII are two famous yeasts developed in Germany 
and used largely in continental Europe because of their ability to ferment 
concentrated solutions quickly. 

14 Among the antiseptics are mercuric chloride, copper sulphate, sul- 
phurous acid, hydrofluoric acid, nitric acid, nitrous acid, salicylic acid, 
formaldehyde, carbon disulphide, chloroform, ether, alcohol, etc. 



274 



The Philippine Journal of Science 



of alcohol, so that the fermentation period is prolonged. Ex- 
tremely prolonged periods of fermentation are to be avoided, 
since solutions are liable to bacterial infection and consequent 
loss of alcohol. On the other hand, dilute solutions cost more for 
distillation and require more space for handling; consequently 
yeasts that will quickly ferment concentrated solutions are to be 
preferred. 

The temperature for fermentation is dependent to some extent 
on the yeast used. Investigation has proved that temperatures 
below 30° C. are usually best. High temperatures are favor- 
able for bacterial growth, especially acid-forming bacteria, and 
the loss of alcohol resulting from evaporation is greater. When 
the yeast causes fermentation, considerable heat is liberated, 
and unless cooling coils are used, the temperature of the fer- 
menting vat rises. In the Philippine Islands, where the usual 
atmospheric temperature is 30° C, the temperature during fer- 
mentation rises as high as 45° C. in some cases. 

EXPERIMENTAL PART 
The solutions of, molasses used were in every case sterilized 
after the sulphuric acid was added. Tests were made on the 
yeast to determine what temperature it is capable of enduring 
without having its powers of reproduction destroyed. Ten per 
cent molasses solutions were sterilized and then inoculated from 
the stock culture of the yeast, with the results recorded in 
Table V. 

Table V. — Effects of heating at various temperatures for ten minutes on 
the yeast found in the fermenting nipa juice. 



Initial treatment. 



Control; sterilized wort; no 

inoculation. 
Control; inoculated; not 

heated. 
Inoculated; heated to40°C. 

Inoculated; heated to 46° C. 



5 (in duplicate) ... Inoculated; heated to 50° C. 

6 (in duplicate) ' Inoculated; heated to55°C. 



7 (in duplicate) .. Inoculated; heated to 60°C. 

8 (in duplicate) _J Inoculated; heated to 65 C C. 



9(in duplicate) .. Inoculated;heatedto70°C. 



Condition at end i 



Sterile Sterile. 



Solid mass of cells. 



Not so numerous as in 2 



Fewer than in 3. Healthy 

condition. 

.—do 

Scattering clusters of 

cells. 

do 

Scattering cells -- 

Condition poor | Good condition 

No cells discernible. Sterile. 



Same as at end of 28 

hours. 
Well developed; 

good condition. 
Do. 



Do. 
Yeast very plentiful. 



Heating for ten minutes has a slight effect at 40° C, hut the 
shock from heating for this length of time is not severe until 



xii, a. 6 Brill and Thurlow: Alcohol from Molasses 275 

a temperature of 65° C. is reached. Yeast heated for ten minutes 
showed the effects of being subjected to this high temperature 
even at the end of the second day. The cells had not become 
so numerous as they were in the other cultures. A temperature 
of 70° C. prolonged for ten minutes destroyed their powers, so 
that they failed to propagate. Heating for longer intervals of 
time at temperatures between 45° and 65° C. will seriously impair 
their powers if it does not destroy them entirely, so that the 
temperature of the fermenting liquid should be down to 40° C. 
before the inoculating solution is added and should be rapidly 
cooled to 30° C. 

Parallel tests of the five cultures of yeast isolated from fer- 
menting nipa juice were made to determine if any difference 
existed in the efficiency of these cultures. Two hundred grams 
of molasses, of the quality shown by the data that follow, were 
dissolved in water, 1 cubic centimeter of concentrated sulphuric 
acid was added, and the whole was given a single heating. Sul- 
phuric and hydrochloric acids have an inhibiting effect on the 
growth of wild yeasts and bacteria ; consequently small quantities 
of these are usually added. After sterilization the contents were 
made up to 1 liter, 0.27 gram of ammonium sulphate was added, 
and the solution was inoculated with 10 cubic centimeters" of 
the inoculating solution. This will be called the standard solu- 
tion for Table VI. The samples were run in duplicate at a tem- 
perature of 30° C. They differed from each other in that the 
second one of each yeast contained 0.06 gram of sodium fluoride 
to the liter. The comparisons were made by removing 100 
cubic centimeters of the ferment, adding 100 cubic centimeters 
of water, and distilling exactly 100 cubic centimeters of the 
mixture and determining the alcohol content by the use of the 
Westphalt balance at 15.6° C. at the end of definite periods of 
time. The acidity determinations were made by titrating 10 
cubic centimeters of the ferment with 0.1 N alkali solution, using 
phenolphthalein as an indicator. It is, therefore, expressed in 
cubic centimeters of 0.1 N alkali necessary to neutralize 10 cubic 
centimeters of the ferment. The alcoholic content is given in 
percentage by volume. The cell count is comparative and is 
the number of millions per cubic centimeter of ferment. In 
Table VI the analysis of the molasses used throughout the labor- 
atory experiments is given. 

The acidity of the ferment could not be determined with ex- 
treme accuracy because of the dark color of the ferment and 
the difficulty of observing color changes. The highest acidity 
corresponds to the lowest alcohol content, showing that the 



276 The Philippine Journal of Science is 

- Table VI. — Data on the molasses used in the laboratory experiments. 





°C. Percent. 






' 






38.1 
21.6 
61.6 
0.2 

0.53 
0.11 



































acidity increases at the expense of the alcohol. The cell count 
at the end of sixteen hours is low in the samples that give a 
low yield of alcohol. The alcohol content is dependent on the 
number of cells. The alcohol content for the five cultures is 
fairly constant. The difference can well be accounted for by 
slight differences in the vigor of the yeast caused by its previous 
experience in the stock-yeast solution. All had the same appear- 
ance under the microscope and reacted identically in the wort, 
which justifies the conclusion that they are the same. The 
maximum yield of alcohol is not reached until the end of the 
fourth day. The results recorded in Table VIII are for the 
standard solutions made up of the same -concentrations of molas- 
ses and sulphuric acid as for Table VII (2 grams per liter solu- 
tion) with the addition of varying amounts of ammonium flu- 
oride. Three of these were kept at an average temperature of 
36° C, while the others were at room temperature (30° C.) 
throughout the experiment. 

Ammonium salts increase the activity of yeast, while fluorides I5 
are reputed to stimulate the yeast and prevent the growth of 
bacteria and wild yeast. 

The data in Table VIII indicate that high temperatures result 
in a loss of alcohol and also that greater action of bacteria takes 
place. The beneficial action of ammonium fluoride is well 
brought out in the higher yields from samples 5, 6, and 8, where 
the larger amounts of this salt were added in comparison with 
7, which had no salt, fermented slowly, and had a comparatively 
low yield. The influence of the fluoride was evident in the free- 
dom from bacteria of samples 5, 6, and 8, and in the presence 
of many bacteria in 7 and somewhat less contamination in 4, 
where only a small amount of the salt was added. The slow 
fermentation is due to the small amount of nitrogen food pre- 
sent. In Table VII, where larger amounts of ammonium salts 

u Backeland, L. H., Journ. Am. Chem. Soc. (1892), 14, 212. 



Brill and Thurlow: Alcohol from Molasses 



277 



o- 
o 

< 


Yield, 
per cent 
of theo- 
retical. 


0> O O U5 O rt rt 




CO CO CO CO CO CO CO 


CCO CO 


si J 

* E § 


o to e© ^ ci oj *- 

(C V ^ N M (O CO 


g s s 


to (d <o <d » « » 


to to to 


i 


•a . 


OO (O Tf o to oi io 


sgg 


CO W CO CO 00 CI c» 


oi eri © 




S3 S3 S3 3 S S J3 


us 


to to to io to to to 


to to US 


5 

o 

J3 


■a . 
'5 >> 

<- 


8.80 
8.76 
8.28 
8.96 
8.88 
9.60 
9.68 


N O O 


J."3 


g??ssss 


ggs 


to to to to to to to 


to to to 


s 


■a . 
o >> 
<- 

O 


8.82 
8.52 
8.52 
8.64 
8.60 
9.00 
9.00 
8.96 
8.64 
9.04 


■* m en <M oo o co 


CO lO N 


8a . 6a ''" 


tj go JO 




5.46 
6.15 
6.12 
6.21 
6.21 
5.91 
6.21 


S o> s 

to' o ■» 



Wjj 00 CO t-' t- 00 00 t- CO CO CO 



SS2SS 



O I ~ o 



cscnt-cocncitotoeooi 



1 

£ i 


V 


OOOUDtOOCO-OUSCO^ 


, XoaJNrih^Nifl'oifl 
t-toc-toc-coc-cototo 




§SSS2S3§2§ 


rlO-HOrtOrHf-I^O 



O «J 3 J! O 2 O 



■«SS|g j 



278 



The Philippine Journal of Science 






OOlO-^OOlOC- 



■EB2 

ci = o 



SS«^ 



ri ■»' 10 m -j 



o oo o o 



to o o o o o 



III 1 



o 8 o o S t ~ 
go dodo cd 



xii, a, 6 Brill and Thurloiv: Alcohol from Molasses 279 

were added, the maximum alcohol content was attained at the 
end of the fourth day. In Table VI the maximum alcohol con- 
tent was not reached at the end of the fourth day in a single 
instance. No effort was made to accustom the yeast to the pres- 
ence of the iiuoride by culture, antecedent to the experiment 
tabulated in Table VIII, and this partly accounts for the slow- 
ness of the fermentation. 

A study of the effect of adding various inorganic salts to the 
ferment was made. The results of this experiment are recorded 
in Table IX. The standard solution used here was 200 grams 
of molasses made up to 1 liter, with the addition of 2 grams of 
sulphuric acid, and then heated to 70° C. After cooling, the 
samples were inoculated with pure yeast culture and the various 
salts were added. 

Table IX shows that magnesium sulphate, sodium chloride, 
sodium fluoride, potassium phosphate, and sea water do not 
stimulate the yeast as do ammonium sulphate and fluoride. Sam- 
ples 2, 3, and 4 gave the largest yields. Samples 3 and 4 had 
two equivalents of ammonium salts added to them, while 2 had 
but one. Samples 1, 5, and 9 had one equivalent of ammonium 
salts added; they excel the remainder of the samples that had 
no ammonium salts added. This is good evidence of the benefit 
of adding ammonium salts. Phosphates 16 stimulate yeast, so 
that its initial activity is increased. They apparently initiate 
the fermentation, that is, if the materials used in fermentation 
could be made absolutely free from phosphates, no fermentation 
would occur. The apparent noneffect caused by the addition of 
the phosphate in samples 8 and 9 is due probably to the presence 
of sufficient quantities of phosphates in the original molasses 
solution to accelerate the fermentation, and the addition of 
further quantities has no apparent effect. 

The ammonium salts keep the ferment comparatively free from 
bacteria, because the growth of the yeast is stimulated and the 
bacteria are crowded out. Ammonium sulphate is as efficient 
for this purpose as is ammonium fluoride and is much cheaper. 
Sea water has a deleterious effect on the ferment. 17 The amount 
of alcohol produced in every case is less than where distilled 
water alone was used. Sample 7, Table VIII (distilled water 
used), gave a yield of 70.5 per cent of the theoretical, while 

" Harden, Arthur, Alcoholic Fermentation. Longmans, Green, and Co., 
39 Paternoster Row, London (1911), 50. 

" The influence of sea water was studied, since salt water from the 
esteros, on which many of the distilleries are located, is often used for 
diluting the molasses. 



280 



The Philippine Journal of Science 



■3 


< 


Yield, per 
cent of 
theoret- 
ical. 

80.6 

89.1 
88.9 

80.6 

70.4 
71.6 
73.4 

80.5 

71.6 

63.4 
62.5 
65.5 


•' Bti ^ <0 <0 «> ° OJ CO ■* O CO t-t£>^H 

all "° 


Condition in 

regard to 
presence of 
bacteria at 
end of 88 
hours. 

Clean 

do 

Fine 

Clean 

do 

do 

Fair 

Clean 

Fair. 

Dirtv 




! 




i 

a 


1 


< 


g s § a s § s s § s s s s 


o_; en co -S 10 n f n t- e h 


a 3 


5-= 


« to <o to (Nt-co t- t- osSn 
id toco to 10 ^ ■*' id tf « w co 


"I 




Ifi CO H i-t IO Ift-^CO . Ifi' ■"* O N M 


3 


U 


b en e 8 -ass 8 s ass 


o_; 




3 
O 


j: loo t-o Oi m n co oj 00 00 u> 

J5 OS C— «-« U3 IO CD <£> "*1* 00 iO CD l£) l« 
" i 




iot- ic ••a 1 w 10 ot£>^ 


1 


"3 


1000 r- 10 » n h 10 as to to 00 




00 i-i 000 000* 


J 

■a 
« 
c 

1 


•0 
■a 

■a 

1 

en 


'5 | '5 '5 '3 g g '3 d w '5 

oj j co « a s « 3g « 
^ ! M M - 1 'S 'S 1 0, g N . 

» : i £ s sgs s 1 5 

3 IS « a " <» rt 0= « 

j 1 0, a "T- rf g " a 
3= ; <b 1 a 5 ? ji ° "" 1 

c !p c c .2 Q a 0" c 

= i 1 I 1 5 3 h E S 3 S 

1 i | 'I 1 £ £ 1 I § '§ 1 

§ i i. is 111 f & Is 

1 ! 1 • i ■ 1 SEE £ | £.: 
bocsbocoocbccsc^tjc^E-gSic 

; d -* 


e 


1 

1 

i 


1 


Sample 
No. 


J .co J J JdJ oi J s j„ 



Brill and Thiuiow: Alcohol from Molasses 



281 



samples 11, 12, and 13, where sea water was used, gave 63.4, 
62.5, and 55.5 per cent of alcohol, respectively. The percentage 
yields range inversely to the concentration of the sea water used 
and are the result of the influence of the impurities carried 
by this water. 

A sample of molasses was prepared as under Table VI and 
exposed to the air without inoculation. At the end of forty 
hours active fermentation had begun, due to inoculation by wild 
yeasts, and the sample showed an alcohol content of 1.40 per 
cent. The maximum alcohol content, 6.61 per cent, was shown 
after one hundred twenty hours had elapsed from the time of 
mixing. 

Much interest is taken in this and other laboratories in the 
properties of the compound occurring in tiqui-tiqui (the polish- 
ings from rice), brewer's yeasts, wheat bran, etc., and which is 
known by the names vitamine, oryzanin,. water soluble B, etc. 
Kurono 1S was led to investigate its effect on yeast in fermenta- 
tion. He made an alcohol extract of rice polishings and added 
small quantities of the dried extract to his ferment. The ex- 
tract accelerated the fermentation even more than does peptone 
or aspaiagine. 

The results obtained by Kurono led us to investigate the effect 
of adding rice polishings directly to the ferment. These results 
are tabulated in Table X. Because of the poor results obtained, 
another sample (No. 7) was inoculated, this time from sample 
6, to determine if the yeast would be more active after having 
become accustomed to the tiqui-tiqui. 

Table X. — The effect of adding varying amounts of tiqui-tiqui (rice polish- 
ing) to fermenting molasses. 



Sample No. 


Tiqui- 
tiqui 
added 
to 1 
liter. 


Alcohol. 


136 hours. 


Alcohol. 


16 
hours. 


40 
hours. 


64 
hours. 


88 
hours. 


112 

hours. 


Alco- 
hol. 


Acid- 
ity. 


Max- 
imum 
yield. 


Yield, 
per 
cent 
of theo- 
retical. 




3- 
5 
10 
20 
«5 
"10 
•20 
5 


0.38 
0.38 
0.38 
0.38 
0.38 
0.38 
0.35 


1.50 
1.50 
1.60 
1.60 


2.50 
2.64 
2.64 
2.50 


3.05 
2.90 
4.13 
3.13 
3.55 


3.99 
3.90 
5.07 
4.05 
4.04 


5.24 
4.66 
5.70 
5.14 
5.31 
5.24 
4.12 


10.86 
11.44 
10 66 
10.84 
9.81 
11.08 
12.32 


5.24 
4.66 
5.70 
5.14 
5.31 
5.24 
4.12 


70. a 
62.7 
76.5 
69.4 
71.3 
70.3 
55.5 


2 








1.54 2.84 
1. 60 2. 98 
1. 60 2. 50 




3.85 1 4.71 
3.41 3.97 















" Cooked. 

' Journ. Coll. Agr., Imp. Univ. Tokyo (1915), 5, 305. 



282 The Philippine Journal of Science »ii 

A comparison of the results with the results obtained when 
the standard solution for Table VIII was fermented shows that 
the addition of tiqui-tiqui has no beneficial influence. One result, 
sample 3, only shows a real increase in alcohol content over the 
sample where no yeast food was added (see sample 7, Table 
VIII). The acidity in every one was high at the end of the 
sixth day, indicating that the addition of tiqui-tiqui contaminates 
the solution or makes the ferment more favorable for the growth 
of bacteria. To extract the tiqui-tiqui with alcohol as done by 
Kurono would be more expensive than using a*mmonium sulphate. 
Culturing the yeast in ferment to which tiqui-tiqui has been 
added did not stimulate its ability to grow in the presence of 
tiqui-tiqui. The use of ammonium salts 10 lowers the yield of 
the higher alcohols (fusel oils) and is, therefore, an advantage 
for this reason. 

Yeast can be invigorated by culture in a nourishing ferment, 
and such yeast acquires a vigor that induces rapid fermentation. 
It can be accustomed to conditions that would ordinarily inhibit 
its grow and that are unfavorable to the growth of bacteria 
and wild yeasts. Advantage has been taken of this property, 
in the method of fermentation known as the Molhant process, 
to increase the resistance of the yeast to more concentrated solu- 
tions of alcohol and to add to its ability to ferment higher 
concentrations of molasses. Mirior, 20 in using this method, 
proceeded as follows: Yeast was added to a small amount of 
molasses of 6° Baume acidified with 3.5 cubic centimeters of 
hydrochloric acid per liter, and the whole was allowed to ferment. 
When fermentation was active, the ferment was pumped to a 
larger tank and more molasses of similar quality was added. 
This was permitted to ferment twenty-four hours and then put 
in a larger tank, where plain molasses of 14° Baume was added 
until the whole mixture was about 12° Baume. Fermentation 
was complete in from twenty-four to thirty hours, and a must 
of 9 to 9.5 per cent alcohol was obtained. The process gives 
60.23 liters of alcohol per 100 kilograms of sugar, calculated 
as sucrose. He states that this is 1.5 liters per 100 kilograms 
more than is obtained by the old process. 

To determine if this process would increase the yields of 
alcohol with the yeast at hand, the following experiments re- 

18 Ehrlich, Paul, Ber. d. deutsch. chem Ges. (1906), 39, 4072; (1907), 40, 
1027. 

"'Bull. Assoc. Chim. de Sucr. et Dist. (1914), 31, 936. 



xii. a, 6 Brill and Thurlow: Alcohol froia Molasses 



283 



corded were planned and performed; solutions were made up 
as described under a, b, and c: 

a. Two hundred fifty cubic centimeters of 20 per cent molasses solution, 
plus 0.25 cubic centimeter concentrated sulphuric, plus 1 equivalent of 
ammonium sulphate were sterilized once by heating and inoculated with 
yeast. Four of these solutions were prepared and allowed to ferment. 

6. Twenty-four hours later 250 cubic centimeters of solutions identical 
with a were added to each of the above four solutions. 

c. Twenty-four hours later 500 cubic centimeters of quantities of solu- 
tions of the strength given below were added to the respective samples: 

1. One hundred twenty grams of molasses in 500 cubic centimeters 
of solution plus the usual proportion of acid and ammonium sulphate. 

2. One hundred thirty grams of molasses in 500 cubic centimeters of solu- 
tion plus the usual proportion of acid and ammonium sulphate. 

3. One hundred forty grams of molasses in 500 cubic centimeters of solu- 
tion plus the usual proportion of acid and ammonium sulphate. 

4. One hundred fifty grams of molasses in 500 cubic centimeters of solu- 
tion plus the usual proportion of acid and ammonium sulphate. 

This makes sample 1 a 22 per cent molasses solution ; 2, a 23 
per cent ; 3, a 24 per cent ; and 4, a 25 per cent. 

The first sample for determination of alcohol content was 
taken at the end of twenty-four hours after the ferment was 
completed, and subsequent samples were taken at intervals of 
twenty-four hours, as recorded in Table XI. 

Table XI. — Yields of alcohol obtained from molasses by the use of the 
Molhant process. 







Alcohol. 


120 hour, jj£ B> 


Alcohol. 


Sample No. 


Mo- 
lasses. 


24 

hours. 


48 
hours. 


72 
hours. 


96 
hours. 


Al- 
cohol. 


Acid- 
ity. 


Al- 
cohol. 


Maxi- 
yield. 


Yield, 
per 
cent 
of theo- 
retical. 




P. cent. 
22 
23 
24 
25 


3.12 
3.05 
3.55 


4.94 
4.94 
4.94 


6.46 
6.70 
6.79 
6.61 


7.16 
7.16 
7.16 
7.81 


7.41 
7.41 
7.89 
8.14 


8.52 
8.40 
8.40 
7.76 


7.25 
7.41 
7.89 
8.00 


7.41 
7.41 
7.89 
8.14 


90.7 
86.8 
88.5 
87.7 






4 









The maximum alcohol content was attained at the end of the 
fifth day. The percentage of alcohol yield of the theoretical 
for sample 1 is the highest obtained in any of the experiments 
carried out and would thus yield a greater revenue to the manu- 
facturer. The highest concentrations gave a somewhat less 
yield of alcohol, but even they are on a par and in many cases 

151034 2 



284 The Philippine Journal of Science iw 

superior in percentage yield of alcohol to those of the preceding 
experiments with more dilute solutions. The length of time for 
the fermentation from beginning to end is two to three days 
longer than by the method in vogue here, but the initial solution 
being of smaller bulk requires less space, and the cost of the 
extra space would be compensated for by the greater yields 
of alcohol obtained. Besides, much space is at present consumed 
by the tardiness in distillation when the fermentation is com- 
plete. It is common practice to allow the must to stand several 
days after fermentation is finished. The distillers do not realize 
that the delay results in a loss of alcohol from evaporation and 
from bacterial action, more especially the latter. 

In order that further data on this method might be collected 
from more thorough trials, the nine experiments recorded in 
Table XII were performed. The nine samples were first made 
by diluting 30 grams of the stock molasses to 200 cubic centi- 
meters and adding the regular amount of sulphuric acid and one 
equivalent of ammonium sulphate. This solution stood two days 
until the fermentation was proceeding strongly, in the case of 
the first of each duplicate, while the second, of each duplicate 
stood for only one day, and then to : 

No. 1 were added 300 cubic centimeters of a solution containing 60 

grams of molasses and the regular amount of acid ammonium 

sulphate. 
No. 2's were added 300 cubic centimeters of a solution containing 70 

grams of molasses and the regular amount of acid and ammonium 

sulphate. 
No. 3's were added 300 cubic centimeters of a solution containing 80 

grams of molasses and the regular amount of acid and ammonium 

sulphate. 
No. 4's were added 300 cubic centimeters of a solution containing 80 

grams of molasses and the regular amount of acid and ammonium 

sulphate. 
No. 5's were added 300 cubic centimeters of a solution containing 90 

grams of molasses and the regular amount of acid and ammonium 

sulphate. 

On the following day the following solution was added to : 

No. 1, 110 grams molasses, regular amount of acid and ammonium 

sulphate in 500 cubic centimeters. 
No. 2's, 120 grams molasses, regular amount of acid and ammonium 

sulphate in 500 cubic centimeters. 
No. 3's, 120 grams molasses, regular amount of acid and ammonium 

sulphate in 500 cubic centimeters. 
No. 4's, 130 grams molasses, regular amount of acid and ammonium 

sulphate in 500 cubic centimeters. 
No. 5's, 130 grams molasses, regular amount of acid and ammonium 

sulphate in 500 cubic centimeters. 



xii, a, 6 Brill and Tina-low: Alcohol from Molasses 



285 



This makes: 

Sample No. 
1 

2's 
3's 
4's 
5's 



One hundred cubic centimeter samples were withdrawn for 
the determination of the alcoholic content at twenty-four hour 
intervals. The results are recorded in Table XII. 

Table XII. — Some further data obtained by use of the Molhant process 
in the fermentation of molasses. 



Sample No. 


Mo- 
lasses. 








Alcohol. 






i 

1 


24 
hours. 


48 
hours. 


72 
hours. 


96 
hours. 


112 
hours. 


144 
hours. 


Maxi- 
yield. 


Yield, j 
per cent j 
of theo- 
retical. 




P. cent. 

20 

22 
23 
24 
25 


3.55 
| 3.74 
1 3.13 
f 3.74 
1 2.89 
1 3.55 
1 3.13 
1 4.05 
1 2.78 


5.79 
5.69 
4.79 
5.99 
4.71 
6.39 
6.01 
5.99 
4.94 


6.22 
6.64 

6.69 
5.61 
6.59 
5.76 
6.95 
5.84 


6.81 
7.34 
6.23 
7.39 
6.15 

6.31 
7.65 
6.46 


6.46 
7.16 
6.95 
7.09 
6.88 
7.68 
7.32 
8.14 
7.42 


6.15 
6.85 
7.41 
6.91 
7.16 
6.91 
7.68 
7.65 
8.06 


6.81 
7.34 
7.41 
7.39 
7.16 

7.68 
8.14 
8.06 


91.1 
89.9 
90.6 
85.7 
83.9 
85.4 
85.4 
87.7 
84.9 













The 20 per cent solutions by the Molhant process have a yield 
equal to the maximum obtained by the regular method. The 
results for high concentrations up to 25 per cent molasses solu- 
tions are uniformly good. In every case they are above 80 per 
cent, the percentage yield that is considered good distillery prac- 
tice. In fact, the average percentage yields are higher than 
85 in each case, and the consistence of the results and the sim- 
plicity of the process make it a valuable method. In the dis- 
tilleries using nipa juice and molasses the custom of mixing nipa 
juice with molasses solution in making up the ferment is followed. 
This is a modified Molhant process, where the nipa juice carries 
the yeast. Such mixtures usually give excellent results in the 
early part of the nipa-juice season, but after the middle of the 
season yields from such mixtures decrease and often become so 
bad that the use of nipa juice must be discontinued. An inves- 
tigation 21 by the Bureau of Science shows that the loss of the 



"Pratt, D. S., et al., This Journal, Sec. A (1913), 8, 377. 



286 



The Philippine Journal of Science 



sugar content of nipa juice is due to the activity of a peroxidase 
elaborated by the flower stalk. This enzyme is not produced by 
long stems to any extent, but as the stems become short, due 
to repeated cutting from the end to renew the flow of juice, 
the quantity of peroxidase elaborated becomes large, and any 
delay in heating or adding sulphites will result in the loss of 
much or all of the sugar present in the original juice. Where 
the juice is mixed with molasses solution much of the sugar of 
the molasses is also destroyed by the enzyme. Sterilization would 
destroy the enzyme, but this would likewise destroy the yeast 
present in the nipa juice and would prevent the inoculation of 
the ferment by means of the nipa juice, which is one of the 
reasons for adding it to the diluted molasses. 

A summary of the results obtained by use of the Molhant 
process is included. 

Table XIII. — Average yield of alcohol by use of the Molhant process. 



Mo- 
lasses. 


Number 
of sam- 
ples run. 




Yield. 




Average, 
per cent 
of theo- 
retical. 


Maxi- 
mum. 


Mi- 
nimum. 


Per cent. 










20 


1 
3 


91.1 
90.4 






90.7 


89.9 


23 


3 


85.5 


86.8 


83.9 


24 


3 


86.4 


88.5 


85.4 


25 


3 


86.8 


87.7 


84.9 



Table XII shows that the more concentrated solutions gave 
high yields of alcohol and that the use of the process will result 
in the saving of fermenting space; fuel for distillation, since 
more concentrated solutions of alcohol will be obtained; and 
the production of dependable high yields of alcohol. 

A sample of ferment, 20 per cent molasses with the regular 
amount of sulphuric acid and 1 equivalent of ammonium sulphate, 
was added and inoculated with yeast, and the temperature was 
kept at 25° C. to demonstrate the effect of lower temperatures. 
The results are given in Table XIV. 

The sample gave a yield of 91.1 per cent of the theoretical 
amount of alcohol obtainable. This result illustrates the ad- 
vantage of keeping the ferment as free as possible from bacterial 
contamination. Temperatures below 30° C, together with vigor- 
pus yeast growth, accomplished this end. 



Brill and Thurlow: Alcohol from Molasses 



287 



Table XIV. — Results obtained for a 20 per cent solution of molasses 
fermented at a temperature of 25". 



Sample No. 


Alcohol. 


hours. 


hours. 
S.69 


H 
hours. 

6.24 


hours. 


m 

hours. 
6.77 


136 
hours. 

6.70 


Maxi- 
mum 
yield. 

6.77 


Yield. 1 
per cent 1 
of theo- 

retical. | 

91.1 


1 . 


0.99 





The next experiment was a repetition of the preceding on a 
larger scale. A 50-gallon barrel was cleaned and fitted with a 
coil, 150 liters of a 22 per cent solution of molasses were placed 
in the barrel, 300 grams of sulphuric acid and 30 grams of am- 
monium sulphate were added, and the whole was heated by pass- 
ing high pressure steam through the coil. Then water was 
passed through the coil, the contents of the barrel were inoculated 
with 2 liters of yeast ferment when the temperature had been 
lowered to 30° C, and the whole well mixed and allowed to 
ferment. The temperature of the solution even during the active 
period of fermentation never rose above 29° C. 

Table XV. — Results of the fermentation of 150 liters of 20 per cent 
molasses solution at a temperature below 30° C. 



Sample No. 


Alcohol. 


16 
hours. 


40 
hours. 


64 
hours. 


88 
hours. 


112 
hours. 


136 
hours. 


Maxi- 
mum 
yield. 


Yield. 

per cent 
of theo- 
retical. 


1 


0.91 


3.34 


4.94 


5.85 


7.16 


7.34 


7.34 


89.8 









The yield of alcohol is well above 80 per cent, the yield of 
good distillery practice. 

With the exception of the preceding test, where an oil barrel 
was used, all the experiments were carried out in glass bottles. 
In order to test our conclusions in a more practical way and to 
meet the possible accusation that such yields would be impossible 
under distillery conditions, permission was obtained to run con- 
trol tests at one of the distilleries of the Islands. 

A vat of about 15,000 liters* capacity was fitted with a coil 
having a surface area of about 7.5 square meters, and connec- 
tions were made by which water from the river or exhaust 
steam from the engines could be circulated through the coil. 



288 



The Philippine Journal of Science 



Three thousand three hundred thirty-five kilograms of molas- 
ses were placed in this vat, and the volume was increased to 
14,000 liters by the addition of warm condenser water. Steam 
was passed through the coil, until the temperature of the ferment 
reached 70° C. Then cooling water having a temperature of 28° 
was passed through the coil. Because of the inadequate size of 
the coil, these operations were extended over a period of more 
than forty hours. The same proportion of concentrated sul- 
phuric acid (2 grams per liter of ferment) was added, but instead 
of one equivalent of ammonium sulphate, two were added (0.4 
gram per liter of ferment), or double the amount used in the 
preceding experiment was used to increase the rate of fermen- 
tation. One hundred fifty liters of inoculating solution were 
then added, and the fermentation was allowed to proceed. 

The molasses employed was different from that used in all of 
the previous work described in this article, as it had been diluted 
to facilitate its removal from the containing tanks. 



Table XVI. 



-Data on molasses used in fermentation experiment tabulated 
in Tables XVIII to XXIV. 



Table. 


Molasses. 


Dry substance in 
molasses. 


Sugar as 
glucose in 
dry sub- 
stance. 


Glucose 
intro- 
duced. 


1 Liters. 

XVIII - - 2,600 

XX - - -| 2,181 

XXII - 2,600 

XXIV 2,700 


Kilos. 
3,335 
3.032 
3,335 
3.676 


Per cent. 
71.0 
72.6 
71.0 
71.0 


Kilos. 

2,201 
2,369 
2,610 


Per cent. 
78.0 
79.8 
78.1 
78.0 


Kilos. 

1,848 
1.756 
1.850 
2.033 



Table XVII. — Data on ferment, results of which are recorded in Table XVIII. 



Water 
Molasses 

Inoculating solution 
Total initial volume 
Final volume 



Liters. 

11,600 

2,500 

150 

14,250 

14,000 



Note. — Sulphuric acid, 2 grams per liter ; ammonium sulphate, 0.4 gram per liter ; initial 
brix, 16.5. 

The yield in Table XVIII is very flattering indeed, and when 
compared with some of the results obtained in the local distil- 
leries, the differences are striking. 

A second control test was run (Table XIX) . 



Brill and Thurlow: Alcohol from Molasses 



289 



Table XVIII. — Results of fermenting molasses on a large scale with 
temperature control. 



Fermen- 
tation. 


Temper- 
ature. 


Cell 
count. 


Alcohol, 

by 
volume. 


Density, 
degree 
brix. 


Alcohol. 

Maxi Yield. 

""^i Per cent 
mum nC .. 
vield of theo_ 
ylem - retical. 


Hours. 
24 
48 
72 
96 
120 
136 


°C. 
30 
33 
30 
31 
29 
28 


10.6 
120.0 
76.0 
71.0 
63.0 
47.0 


2.98 
6.91 
7.01 
7.25 
7.16 
6.92 


•12.2 
6.4 
4.2 
3.9 
3.8 
3.8 














7.25 


89.8 











' Original density, 16.5 brix. 

Table XIX. — Data on ferment, results of which are recorded in 
Table XX. 

Liters. 

Water 12,021 

Molasses 2,181 

Inoculating solution 150 

Total initial volume 14,350 

Total final volume 14,150 

Note. — Sulphuric acid, 2 grams per liter ; ammonium sulphate, 0.4 gram per liter ; initial 
brix, 16.5. 

Table XX. — Results of second trial of fermenting molasses with temper- 
ature control. 



Fermen- 
tation. 


Temper- 
ature. 


Cell 
count. 


Alcohol. 

by 
volume. 


Density. 


Alcohol. 


Maxi- 
mum 
yield. 


Yield, 

per cent 
of theo- 
retical. 


Hours. 
16 
40 

s 

112 


°C. 
33 
33 

31.5 
31 
30 


35 
90 
80 
76 
72 




16.3 
9.7 
4.9 
3.8 
3.7 






3.84 
6.27 
6.81 










6.81 


90.7? 







This result is very slightly superior to that of Table XVIII 
and shows that consistent yields can be procured by exercising 
the precautions used in these two experiments. 

A control test in an adjoining vat was run to determine the 
yield under present distillery methods. This vat had no cooling 
coil, and the ferment was not inoculated. 



290 



The Philippine Journal of Science 



1917 



Table XXI. — Data on ferment, results of which are recorded in Table XXII. 

Liters. 

Water 12,600 

Molasses 2,500 

Total initial volume 15,100 

Total final volume 14,950 



Note. — Sulphuric acid, 2 grams per liter; ammonium sulphate, 
brix, 16.2. 



i.4 grram per liter; initial 



Table XXII. 



-Results obtained by fermentation of molasses without inocu- 
lation and without cooling coil. 



Fermen- 
tation. 








Density. 


Alcohol. 


Temper- j Cell 
ature. ' count. 


Alcohol, 

by 
volume. 


Maxi- 
mum 
yield. 


Yield, 
per cent 
of theo- 
retical. 


Hours. 
16 
40 
64 
88 
112 


°c. i 

31 j 1.3 
38 j 78.7 












' 






6.16 1 6.8 
6.24 6.4 






84 
32 


47.9 
41.0 


6.24 


70.1 











The yield recorded in Table XXII is only 77.9 per cent of what 
the yield was in Table XVIII. This excess of Table XVIII over 
Table XXII can be ascribed to the one sterilization at the begin- 
ning, to the use of pure yeast cultures, and to the reduced tem- 
perature during fermentation. The high temperature in the 
latter case is conducive to bacterial growth, greater evaporation 
of the alcohol, increase of acidity, and incomplete fermentation, 
because of the decreased production of enzymes from the yeast. 
The attenuation, which is an approximate measure of the degree 
of alcoholic fermentation, was not as complete in Table XXII, 
where it proceeded from 16.2 to 6.4, as it was in Table XVIII, 
where it went from 16.5 to 3.8. 

As has been already mentioned, many of the distilleries of 
the Philippine Islands ferment nipa juice either alone or mixed 
with diluted molasses. In order that the efficiency of this method 
might be determined, the experiment tabulated in Table XXIII 
was carried out. 

Table XXIII. — Constitution of ferment, results of which are recorded in 
Table XXIV. 

Liters. 

Water 10,000 

Tuba 8,700 

Molasses 2,700 

Total initial volume 21,400 

Total final volume 21,000 

Note.— Initial acidity, 8.5 cubic centimeters 0.1 N alkali in 10 cubic centimeters of ferment ; 
Initial brix, 12.6. 



Brill and Thurlow: Alcohol from Molasses 



291 



Table XXIV. — Results for a mixture of tuba and molasses. 



Fermen- 
tation. 

- 

Hours. 
16 
40 
64 
88 
112 


Temper- 
ature. 

42 
40 
37 
SS 
33 


Cell 
count. 

74 
54 
50 
48 
43 


Alcohol 

by 
volume. 

5.09 
6.61 
5.64 
5.46 


Density. 

7.4 
5.9 
5.8 
5.7 


Acidity. 

10 
10 

11 

14.5 
16.5 


Alcohol. 

1 


Maxi- 
mum 
yield. 


Yield. 
pi;r cent 
of theo- 
retical. ' 


5.61 


•66.7 


















« Assuming that the sugar content is 10 per cent, which is leas than good tuba should 
contain. 

The yield recorded in Table XXIV was somewhat less than in 
the preceding test and is not comparable with the yield ob- 
tained in fermenting a solution in which the temperature is kept 
below 30° C. and vigorous clean yeast is used for inoculation. 
The acidity of this solution increased rapidly at first, doubtless 
owing to the activity of the peroxidase carried in the tuba. The 
activity of the peroxidase will decline when the alcohol content 
of the ferment increases, and this period is marked by the 
period of no change in the acidity of the ferment, but after a 
short interval it again increases, due to the activity of bacteria. 

The density, as shown in Tables XXII and XXIV, has not 
decreased to the same extent as in Tables XVIII and XX. Many 
results better than these recorded in Tables XXII and XXIV 
are obtained by the distilleries of the Islands. The distillery 
to which we had access has records of better yields than these, 
but the fact that such yields occur when the "tuba has gone 
bad," or because of other reasons, shows the need for closer 
supervision and greater knowledge of the best conditions for 
fermenting this molasses. 

RECOMMENDATIONS 

We recommend the sterilization of the molasses solution 
wherever this can be attained without too great cost or the 
installation of extra machinery. If this is impossible, the use 
of good water for diluting the molasses to a definite density, 
about 16.5 brix, is essential. Two grams of sulphuric acid and 
at least 0.4 gram of ammonium sulphate to every liter of ferment 
should be added. Inoculate the ferment with clean yeast. One 
part of fermenting wort to 100 or 150 parts of ferment is the 
right proportion. Seed for the production of this inoculation 



292 The Philippine Journal of Science 

yeast can be obtained from the Bureau of Science, and at the 
same time instruction can be procured for keeping this com- 
paratively free from infection. Clean yeast can be obtained by 
inoculating a small quantity of sterilized 10 per cent molasses 
wort by means of a sterile platinum wire or by the addition of 
a few drops of stock yeast. When this is fermenting strongly, 
it should be added to a larger volume of sterile molasses solution 
(12 brix), preserving the proportion of 1 to 100 (or 150), and 
the operation can be repeated with this volume until the content 
of yeast is sufficient to inoculate the vat, which is ready to be 
fermented. If care is exercised in choosing ferment in which 
the yeast is not badly contaminated, the new vats can be in- 
oculated by the addition of stock solution from vats in active 
fermentation. By beginning with a 10 per cent solution of 
molasses and making each successive solution slightly more con- 
centrated until a brix of 16.5 is attained, the ability of the yeast 
to ferment in more highly concentrated solutions is strengthened. 
In other words, the principles embodied in the Molhant process 
are being put in operation. The point of maximum alcohol 
content should be determined by frequent determinations of the 
alcohol present in the fermenting solution. When this point is 
reached, the ferment should be distilled without unnecessary 
delay. The ebullioscope is used by some concerns for this deter- 
mination, but its use is unsatisfactory and cannot be relied upon. 
The manufacturers specifically say it is for use in determining 
the alcohol content of dilute alcohol solutions, dry wines, etc. 
Where sugar or other solid is present in solution, the results 
obtained by it are unreliable; consequently the alcohol content 
of fermenting molasses cannot be measured by this instrument. 
The temperature of the ferment should be kept between 28° 
and 30° C. by suitable cooling coils. 

SUMMARY 

Some statistics of the alcohol industry in the Philippine Islands 
and abroad are given. 

The data of a number of experiments with fermenting molas- 
ses are recorded. These show the influence of various salts on 
the rate of fermentation, the yield of alcohol, and the virility 
of the yeast. They include several trial experiments on a large 
scale with and without temperature control. 

Certain recommendations are made for the improvement of 
the present methods of fermenting molasses. 



THE RADIOACTIVITY OF THE WATERS OF THE MOUNTAINOUS 
REGION OF NORTHERN LUZON ' 

By George W. Heise 

{From the Laboratory of General, Inorganic, and Physical Chemistry, 
Bureau of Science, Manila) 

ONE PLATE AND TWO TEXT FIGURES 

In a previous paper 2 the results of a series of measurements 
of the radioactivity of typical Philippine waters were reported. 
About 90 waters were examined, including 44 from springs in 
the mountainous region of northern Luzon. This province is, 
in the main, a volcanic region, of which the — 

Geology clearly indicates, geologically speaking, the evidence of only 
recent vulcanism.* 

It contains an abundance and a variety of springs, many of 
them hot and heavily mineralized; hence it is an especially 
desirable place for work of this kind. 

In the work referred to, there was no apparent relation to 
be deduced between the radioactivity of the waters and the 
chemical quality of the geological formations from which they 
were derived. The work done was admittedly preliminary in 
character, and the result could not be considered conclusive. 

With the exception of certain sources in Ifugao subprovince, 
the springs examined in Mountain Province were uniformly low 
in radium emanation content. There was no apparent reason 
for this peculiarity, as the waters studied were from many 
different geological formations and showed great variation in 
chemical quality. 

It seemed of possible significance that the only strongly radio- 
active waters were encountered in Kiangan and Banaue, the 
only places east of Polis Range at which we had made examina- 
tions. However, the available time and the exigencies of moun- 
tain travel did not permit more extensive investigation. Re- 
cently it has been found possible to extend the work, with the 
results set forth in this paper. 

In addition to the places mentioned in the preceding report, 
the accompanying map (fig. 1) shows the following places at 

1 Received for publication July, 1917. 

2 Wright, J. R., and Heise, G. W., The radioactivity of Philippine waters, 
This Journal, Sec. A (1917), 12, 145. 

' Eveland, A. J., Notes on the geology and geography of the Baguio 
mineral district, ibid. (1907), 2, 207-233. 

293 



294 



The Philippine Journal of Science 



which work was done this year: Noso, Aritao, Bambang, Ba- 
yombong, Solano, Bagabag, Salinas, and San Luis, Nueve Viz- 
caya; Amdangle, Banaue, Sapao, Aoua, and Monhuyhuy, in 
Ifugao, Mountain Province; and Loo, Buguias, Lutab, Daklan 
Ambuklao, and Baguio in Benguet, Mountain Province. 

In addition to the regular equipment used for the determina- 




Fig. 1. A part of northern Luzon. 

tion of radioactivity, we took with us a small chemical field 
laboratory, modified to meet the unusually severe conditions of 
mountain travel. The use of this laboratory not only enabled 
us to obtain data of interest in the study of radioactivity, but 
it was also of definite value in furthering the field survey 4 of 
Philippine water supplies now carried on by the Bureau of 
Science. 



4 Heise, G. W., Water supplies in the Philippine Islands, II, ibid. (1915), 
10, 137. 



Heise: Radioactivity of Waters, Northern Luzon 295 



Except that a Schocky- Willis radioscope was used instead of 
the Spindler-Hoyer electroscope, apparatus and method were 
essentially the same as those previously described in detail." 
Because of the great bulk of the Schocky-Willis apparatus and 
its complete lack of accessories for field work, a smaller brass 
ionization chamber was constructed in the Bureau of Science. 
The collecting cans and other necessary field equipment from 
the Spindler and Hoyer outfit were added before the apparatus 
was taken into the field. By cutting down the size of the ioniza- 
tion chamber, the sensitiveness was reduced approximately to 
40 per cent that of the Spindler and Hoyer electroscope. This 
lack of sensitiveness, together with the mediocrity of the tele- 
scope and the difficulty in the field of getting readings with 
the leaf charged to the same potential, made readings somewhat 
uncertain and increased the percentage of error. Check deter- 
minations made it appear probable, however, that the limits of 
error mentioned previously had not been greatly exceeded. The 
results obtained may be, therefore, considered reliable, except 
for waters of low activity, for which the exact value of the 
emanation content is of minor significance. 

The modified apparatus, as assembled in the field for a deter- 
mination of radioactivity, is shown in Plate I. 

As before, no attempt was made to determine anything but 
radium emanation content. The data secured for the radio- 
activity of the waters of northern Luzon are shown in Table I, 
the results obtained in 1916 and already reported 5 being in- 
cluded to enable the presentation of the work as a whole. 
Table I. — Radioactivity of waters of northern Luzon. 



Location (province, sub- 
province, and town or 
district). 



I__ I 1* 



May 31,1916 



MOUNTAIN PROVINCE. 

Amburayan, Tagudin . 



do 

do 

May 24,1917 
— .do 



do 

do 

Benguet, Ambuklao. 
do 



jRadium 
Eleva- \ emana- 
tion Irp tionper 

■"ovelS^Hter.. 

sea 1< "- U1C - (rrams 

level. RaX 

10-«. 



Meters. °C. 



?i:2 



341 



■10 



j trace 

I 137 
trace 



Flowing well near ce- 
metery. 

Flowing well, plaza 

Flowing well near school 

North spring 

; West spring 

May 1, 1916 I Benguet, Baguio J Antimok River spring* . 

a Nonthermal. 

* Waters which, owing to the nature of the source, could not be obtained just as they 
emerged from the ground, or those which may not have been typical ground waters owing 
to their derivation from springs believed to be local or temporary. 



(*) 



' Wright and Heise, loc. cit. 



296 The Philippine Journal of Science 1917 

Table I. — Radioactivity of waters of northern Luzon — Continued. 



10... 



16. 



19. 



21... 



37. 



April 27. 1916 
May 27,1916 



May 26.1916 
April 28. 1916 



April 26. 1916 
May 

May 



, 1!'17 



1916 

April 27. 1916 
April 28. 1916 
May 28.1916 

May 19,1917 

do 

May 20.1917 

do 

May 23.1917 

do 

do 

do 

May 5, 1916 

do 

Dec. 1916 
May 29,1916 
May 22,1916 
Dec, 1916 
May 21,1917 

do 

do 

May 4. 1916 

do 

May 7, 1916 



Location (province, sub- 
province, and town or i 
district). 



Benguet, Baguio Camp John Hay. Carino 



do 



do 



Eleva- 
tion 
above 



Meters. 
?1,300 



ice- 



Camp John Hay, 

plant spring. 
City spring, Bokawkan 
j Road. 
do ! Consolidated Mine I 



! 1.400 



.do 



1,350 



1,350 j 



1,350 



1.150 



..ion 



1,400 

1,400 

1,400 

1.400 
1,376 
900 
1,360 
?1,300 
1,100 



spring. 

.do I Dominican Home spring 

.do Federle's SprinK. Suyoc 

Trail. 
Government Center 
spring. 

.do Small spring, adjacent to 

No. 13. 
Headwaters Mine 

spring. * 
Sanitary Camp spring... 

Pakdal spring 

Spring at slide, Bua 
Road. ' 

Benguet, Buguias Salt spring near presi- 

dencia. 

.do Salt spring, south of 

presidencia. 

do Warm spring across river 

do Warm spring near No. 21 

Benguet, Daklan Asin (salt) spring 

do Badukbuk solfatara 

do Kakomotan spring 

do Palungod spring 

Benguet. Haight's i Spring near house' j 2,550 I 

do : Spring near barn*.. '■ 2,650 

Benguet, Itogan ! Itogon hot spring ' ?1, 150 

Benguet, Klondike j Hotspring 180 j 

Benguet, Loo. _ _ Rest house spring I 1,800 

Benguet, Lubong. Hotspring ?50 

Benguet, Lutab ___| Duakan spring i 1,200 

do... _ ! Roadside spring, Kaba- | 1,150 

yan. 

do | Ubud spring. I 1,100 

Benguet, Mountain | Spring near rest house*.. 1,770 
Trail. Kilometer 30. j 

do j Spring on Atols Trail *... 1,750 

Benguet, Mountain | Spring near rest house*.. 1 2.200 
Trail, Kilometer 80. I 
" Nonthermal. 

b Emanation was high ; sample of gas bubbling through a pool of hot water. 
* Water which, owing to the nature of the source, could not be obtained 
emerged from the ground, or those which may not have been typical ground 
to their derivation from springs believed to be local or temporary. 



Tempe- 
rature. 


Radium 
emana- 
tion per 
liter as 
grams 
RaX 
10-". 


°C. 




„ 


194 


„ 


122 


,, 


107 


23 


93 


20 


111 


w 


137 


22 


163 


22 


200 


21 





(«) 


29 


w 


trace 


23 


70 


60 





57 





48 





45 


trace 


52 


95 


770 


(») 


22 





25 


270 


w 





(•) 





?60 





?55 


trace 


w 


trace 


7200 





22 





20 


180 


21 


140 


(•) 





w 





1 






just as they 
waters owing 



xii, a. 6 Heise: Radioactivity of Waters, Northern Luzon 297 
Table I. — Radioactivity of waters of northern Luzon — Continued. 



I Location (province, sub- 
province, and town or 



Dec. 

I May 14, 

| do. 

May 13, 
May 15. 
May 11, 

do. 

May 12, 

do. 

May 9. 
May 10. 
May 9. 
May 20. 

do_ 

May 17, 
May 16, 

do. 

May 17. 
May 12, 
May 13, 
.....do. 
May 18. 
May 12. 
May 19. 
May 19, 
May 17, 
May 11. 
May 17, 
May 8. 



1916 Benguet. Trinidad 
1916 Bontoc. Bontoc—. 

— .do 

1916 do _ 

1916 I Bontoc, Mainit _ — 
1916 ! Bontoc, Sagada... . 



id,; 



.do., 
-do.. 



.do 



1916 | Lepanto. Cervantes 



—.do. 

.—do.. 

Ifugao, Aoua 

-—do 

Ifugao, Banaue 

— .do 

— I—* — 

1916 do ._ 

1917 ; Ifugao, Kiangan 

1917 do 

.1 do 

1916 I do 

1917 do 

1916 ' do 

1917 do 

—do.. 

—do 

Ifugao, Sapao 

Lepanto, Mancayan . 



Eleva 

al i o°v n e Te r e - 

sea ralurc 
level. 



May 5, 
May 6. 



May 7, 
May 1. 
May 9. 
May 4, 
May 8, 
May 3, 
May 6. 
do. 



Spring near municipio— 

City spring 

Spring near city stables. 
2 springs, barrio Samoki 
Mainit (hot) salt spring.. 

Mission spring — 

Underground river I 

Spring, barrio Tetepan...; 
Spring at slide, Tetepan _. 

Municipal supply j 

Hot spring, river bank ... 

Comilias hot spring I 

Spring at Kilometer 65 — | 
Spring at Kilometer 76 

Bognakan spring 

Kiakop spring 

Magabon spring 

Paypayan spring 

Adhang spring 

Adukpung spring 

Adungbu spring 

Atuda spring 

Lakdo spring * ._ 

Malpao springs (2) 

Lubungan spring 

Piko spring* 

Piko spring 

Bolanan spring 

Spring. Balili trail 



Meiers. 

1,250 

920 

920 

920 

1,300 

1,650 

1,450 

1,750 | 

1,300 ' 

71.000 



770 
500 
1,300 
1,300 
1,150 
1,300 
1,150 
1,150 
1,000 
850 



Bagabag 

—.do 

Bambang — 
Bayombong- 

do 

Boscaran ... 

Noso 

Orcoring 

Salinas 

San Luis 

Santa Fe 

Solano 

do 



Bafios spring.- _ 

Small pump well 

do _■__ 

Bangan spring 

Plaza pump well 

Spring near school 

Spring near rest house- 
Spring near road 

Salinas salt spring 

Spring in quarry 

Santa Fe spring 

Solano spring 

Solano pump well 



550 
(?) 



Radium! 
emana-J 
tion per: 
liter as 
grams 
RaX 
10-12. 



oC. 
"21 

w 
(») 

w 

100 
« 
(•) 

w 
w 
w 



27 




trace 
trace 





111 

trace 





1,200 
1,325 

720 
189 
190 
150 
945 
1,058 
900 
175 
114 



27 


130 


27 


190 


27 


130 


26 


130 


26 


325 


24 





30 


130 


31 


95 



240 



a Nonthermal. 

* Water which, owing to the nature of the source, could not be obtained just as they 
emerged from the ground, or those which may not have been typical ground waters owing 
to their derivation from springs believed to be local or temporary. 



298 The Philippine Journal of Science im 

The table is self-explanatory, for the most part, but there are 
a few points in connection with it that merit attention. 

The springs in the list are typical and are representative of 
the country in which the work was done. They comprise the 
prominent and better known sources along the route of travel, 
among them several used for salt manufacture (Salinas, Mainit) 
and several hot mineral springs with reputations for medicinal 
virtues (Itogon, Mainit, Comilias). 

The highest activity recorded (1,325 X 1(H 2 ) was found in 
Adukpung spring, in Kiangan. 

It will be noted that Piko spring, in Kiangan, was examined 
both in 1916 and 1917 and that there is a marked discrepancy 
between the two determinations. This difference does not neces- 
sarily indicate a variation in emanation content, nor even a 
serious error in the determinations, as the water from the spring 
in question flows into a small, covered reservoir before it is 
allowed to emerge ; therefore a sample cannot be secured directly 
at the point of emergence. 

As might be expected in a volcanic region, there are a number 
of solfataras in the area under discussion, notably at Bolotoc, 
at Daklan, and near Monhuyhuy. The one at Daklan, which 
we visited this year, is characterized by a number of vents, from 
which hot gases and vapors emerge, principally hydrogen sul- 
phide, sulphur dioxide, and steam. We were unable to find 
any springs at this place. There were a number of excavations 
in which surface run-off and perhaps condensed steam had col- 
lected, forming great caldrons of hot water, some of them used 
as baths through which the gases and vapors bubbled. There 
were also a number of hot mud "springs," where the gases broke 
through semiliquid mud with the peculiar sound from which the 
solfatara presumably has derived its Igorot names (Barutbarut, 
Badukbuk). As the waters in the excavations just mentioned 
were probably surface run-off, and as they were constantly 
aerated by the gases bubbling through them, their activity was 

6 This spring is peculiar. It is located in the wall of a rice paddy and 
emerges at a level only a few decimeters below that of the water in the field. 
As it is separated from the rice-paddy water by less than 5 decimeters of 
earth, this spring looks like a mere seepage. We were assured, however, 
that it is a true spring, whose flow does not fail throughout the year, even 
during the months when the rice field is quite dry. Analysis of the water 
in the rice paddy showed marked differences from that from the spring, and 
the determination of activity seems to furnish further proof that the source 
is a real spring. 



xii, a, 6 Heise: Radioactivity of Waters, Northern Luzon 299 

of minor interest. A collecting can, full of the gas taken from 
one of the excavations by downward displacement of water, 
proved to be radioactive, as indicated in Table I. 

For completeness, I have compiled the available chemical data 
on the waters studied and have included them in Table II. All 
of the waters tested for radioactivity in 1917 were analyzed at 
the source by Mr. A. S. Behrman, chemist of this Bureau, who ac- 
companied me on the field trip. In the case of highly mineral- 
ized waters, or when additional gravimetric determinations were 
desired, samples were taken to Manila in glass-stoppered bottles 
and analyzed in the laboratory. It should be noted that the data 
given under "total hardness" were obtained by a modification 
of the Blacher ' potassium palmitate method and represent com- 
bined calcium and magnesium content. 

Of the numerous salt springs examined, none showed more 
than small amounts of activity. The hot springs, too, were only 
slightly or not at all radioactive. These results are in agreement 
with those obtained elsewhere 8 and are to be expected from 
the low solubility of radium emanation in hot water or con- 
centrated salt solutions. A number of waters low in radio- 
activity are also low in dissolved mineral matter. This is prob- 
ably of little significance, except to show that certain sources, 
instead of being deep-seated springs, were but little more than 
seepage water, percolating the soil for comparatively short dis- 
tances. It might be pointed out that the most radioactive waters 
encountered were high in calcium and magnesium content, in- 
dicating an origin in calcareous material. The work done in 
the Philippines up to the present time is still insufficient to 
justify conclusions, so that the cases noted should be regarded 
as isolated observations, at least for the present. However, it 
is worthy of note that this peculiarity is distinctly at variance 
with the usual observation 9 that the water from igneous rocks 

' Blacher C, Griinberg, P., and Kissa, M., Die Verwendung von Ka- 
lhunpalmitat bei der Wasseranalyse, Chem. Zeitg. (1913), 73, 56-8. This 
method has been adapted to field work by Mr. A. S. Behrman and now 
forms one of the regular routine determinations in the Bureau of Science 
field assay of water. 

8 cf . von Hofer, H., Radioactive springs, Intern. Zeitschr. Wasser- 
versorgung (1914), 1, 52-5, 90-3; through Chem. Abst. (1915), 9, 1714. 

9 cf. Clarke, F. W., Data of geochemistry, Bull. U. S. Geol. Surv. (1916), 
616, 315. Sahlbom, N., Arkiv. Kemi. Min. Geol. (1915), 6, No. 3, 1-52; 
through Chem. Abst. (1916), 10, 1134. 

151034 3 



300 



The Philippine Journal of Science 























» "2 






o o 




o o 


. o o 




' o 


I o o o o o o 


■M8 
A 5 






s gg 




!5»» 

i A 


; s g 
jv" 




i d 
i V 


10 *o \a © «o u: 






00 N O O CM O OOOOO 


1#3 




_ « ri « • « • 8 «• « » 8 8 g d ^ ^ oo rf „ • 

: H cot ~ § aS § « rtrtco a a a £ £ S "^ £ ^ « 


«-; 




OlOWOOTrcgNOi-tOOOi-HONCOO 


. o o 


.JicO 




coooO'O'cocoeo-^tt'waoooooNoooo 


1 §3 


CQOg 








§J5 




•9 


S "3 




B E E E E B 6 E E B "E 'S B B g E 6 SB 6 


UaZ 










OlQOOOOCOtOCO<0u3*-l 


OUSOIOOOOkOUS 


18 




l NN v loto ' N ° , vv eo 


" N 1^8|^ 








id -w 


o 








£ 




3 5 ,»S S2 «» S c- ooSS 








£ 












.Ems 




OOOWOOO O O U3 lO 00 O lOOOOO 


P 




sSiS^sssa^^'SsVii-s 


<^ 




<N 


• -; 




ooooooooooooocooooo 


o o 


a£0 




g^-ssss^^sa^sss 


s s 


<! o 




* 














j 




U30©0©00000©010©0© 


o in 


£ O 3 

<3 s 


1 

3 




AA A A A "1 














b 

c 


























'b 






























































a 

IE 




t 




bo ' 








Ifl 








i 

o 




« t 

11 

1 1 


6 

1 

( 


1 
>> 

^ 

o Q 


bo ft 

B «> 

11 ! 

Ill 

O G fc 


V o g> 

1 11 b 

E .* 2 ■- 

2 •3 ■ s c 
E « a « 
E g-| « 
! "£ ' 3 -3 

O W K (X 


| a 

a m 
™ 

E S 

8 « ' 

||] 

a a, "o 

WWW 


11 
bo bo 

fill 

g 1 1 

a a ■= 
& * « 


(X 


■I «■ 

!'l 

a m 

J ! 

s3 £ 


e 

h 


a 
g 












































■g'C 












































E o 


6 

z 
3 


11 

a 3 -3 

J J 


o 

I 






















a 

'3 

i. 




'5 

3 


c 








32 


2 


§ < 


P3 




















PQ 


CQ Q 






l« 


2 


If<3 


5 Q 


c 

a 


c 

a 


Q 


c 

p 


c 

c 


c 

C 


t 

a 


c 


o 
C 


bo O 


a o . _ 
3 = o 
bo bo O Q 


o 

Q 


§ 






J 




< w co ca co cc 



if 



E -a 
. 3 

oJ )S o 

g 35 

too" «* 
e . u 

E io S 

■- ~Z ™ 

Hi 
?j: 

e* 6 2 

S 5 " 



11* 
3 S e» 

&S8 



xii. a. * Heise: Radioactivity of Waters, Northern Luzon 301 



OOOOOOOOOOCOOOO =00000 

g S IS I' g g g g S' S '4 I I 3 § g g I -" I S 



^ '5 m « '£ '£ '£ 



3 S 5 



? § ?2 I ?? g I S 2 3 S 3 S g 



410.0 
21.3 
95.0 

106.0 
26.8 
40.3 


59.8 
122.0 
128.0 
146.0 
115.0 
195.0 
226.0 
280.0 
170.8 
280.0 

45.0 
238.0 
305.0 

61.0 

70.0 


238.0 

195.0 

170.8 

nil 

159.0 

293.0 


55 


b> 




a c c c c 




c c c c c c 



o o 

1 1 



$9* 



t-' ci ^ ^ eJ co ci ei I o£ ej ej ^ g 



s s s 

o - o" d 



S-ojj 



ooooooooo © ■« O W >-l -H 



OOOO OOOOIO 






a 



ill 

SHE 



oooooooooOooo 
©<NOtoooco'*cococr:LnLQic 



S S S « 3 S I s 



:6b 



9.SE 



a ° S i " " '5 S I b 
,3 § ft .2 ■{§ § "° ,9 - 3 



S 8 : 8 : § " ~ 



is & 

C ft ft 

•s S £ 



|Z! -S 2 

"3 S 



l-JJ 

CO C .S 






3"OO0OOOOOOOOO 

to -S ? « p Q Q <?. QQOQQQ 



; bo ; ; 

Si i S 



si Si 



111 



&z 



3 sfi^Sfi 



302 



The Philippine Journal of Science 



a -3 
£ SO 
■2-Su 

a o 




c 




c 


i 


C 
1 


o 

1 




"3 ao 


= 2 ° 2 2 "" N 


111 

m.83 


ooooooo 

3 1 1 1 s s i 


ijo 

«J °i § 


s g K a a a a 

c c c c c c 


is 


10 o o 10 m co eo 


2 

1 


S3 <o co 5 S S3 ?2 

o' o t-i o o d d 


1 =o 


t- o o . o a 

» = § a ^ s § 


Alkali- 
nity 
(CaCOs). 

100.0 
340.0 
3, 700. 
210.0 
130.0 
160.0 
160.0 


1 j? « 1 »• 2 2 § IS 8 § 

III i 1 
6 s * 




1- 




T 


c 

V 


1 
G 


1 


c 

T 


1 

p 

1 

ft 




Location (province, subprovince, 
and town or district). 

1 


o 
o 

0. 

<! 
< 

< 
> 

111 


c 
C 


c 


I 

t 

a 


2 

5 

H 


i 

a 


o 


! 



-S w i 



g 

g 

3 

8 

c 
M 

s 
s ° 

. CM 

li 

il 

"3 co 

I 8 . 
.. o 

- s 



a 2 

© S 



xii, a. 6 Hcise: Radioactivity of Waters, Northern Luzon 303 

is generally more radioactive than that from sedimentary 
deposits."' 

The geology of the area discussed in this report has unfortun- 
ately not been worked out in sufficient detail to justify generaliza- 
tions; hence it is not feasible at this time to attempt to deduce 
any definite relations between radioactivity and the geology of 
the water-bearing strata. The following description by Smith n 
of a typical portion of this region is appended for the sake of 
completeness : 

This region is typically mountainous, and from the character of the relief 
we must consider it as being in the stage of "topographic youth." * * * 

There are two distinct types of topography in the region covered by 
this paper, and these are directly to be attributed to the character of the 
geologic formations. In the country to the west of the Polis Range the 
formations are mainly volcanic, and we find there an irregular, rugged, 
accentuated relief. The elevations vary from 370 meters to 2,400 meters 
or more. 

East of this range the formations are folded sediments, giving rise to 
a more regular topography, and in places the hills and mountains are 
nothing more than tilted blocks of sandstone. On the eastern slopes these 
present long, gentle inclines, but to the west they form steep escarpments 
with here and there a saw toothed skyline. As one goes farther to the 
east, approaching the valley of the Cagayan, the mountains become mere 
foothills. * * * 

A cross section from the west coast at Tagudin northeastward to the 
edge of Cagayan Valley gives as good a general idea of the formations and 
structure of the region as one could expect to get by any procedure short 
of a detailed survey of the whole country. Fig. 2 is a graphic attempt to 
record my interpretation of the main facts. 

Near the west coast we find gently folded shales and sandstone whose 
inclination increases as we go toward the Malaya Range, being much 
contorted as we get well into the canon. The Malaya Range is essentially 
a mass of porphyry or, to be more exact, andesite. * * * 

The town of Cervantes is situated on a small tongue of high ground 
between the Abra and one of its branches. The underlying rocks are 
practically the same as those found in the highland on both sides of the 
town. As we go toward Bontoc we find the same andesitic mass with, 
however, several large outcroppings of quartz. 

When we reach Bagnan and Sagada we find tuffs and reef limestones 
overlying this igneous mass 

"From Sagada to Bontoc extrusive rocks, almost entirely andesites and 
dacites, are encountered but east of Bontoc the formations become very 
shortly diorite and granite. This belt of granitic rocks is only from 8 to 
10 kilometers wide. There is more andesite, or rather a very fine-grained, 

10 One of the most radioactive waters in the lowlands, from Sibul Springs, 
Bulacan, is also in a calcareous formation, and others high in activity issue 
from tuffs and conglomerated igneous materials. 

u Smith, W. D., Notes on a geologic reconnaissance of Mountain Province, 
Luzon, P. I., This Journal, Sec. A (1915), 10, 177-209. 



304 



The Philippine Journal of Science 



almost aphanitic phase of diorite, succeeding 
the granite on the east as far as the town 
of Barlig, where steeply dipping sandy 
shales are encountered. These incline to 
the southeast. 

Sandstone makes up the main mass of 
Mount Amuyao, and from there east to Ca- 
gayan Valley sediments occur. In places, 
as at Natonin, there are small areas of 
extrusive rocks overlying the Tertiary 

j sediments. 

3 

8 A comparison of -the radioactivity 

| of the waters of the mountains of 
| northern Luzon and that of the 
| waters examined in the Luzon low- 
s lands indicates that the typical 
I lowland waters show the higher 
" radioactivity. The highest radio- 
s activity (1325 X 10- 12 ) determined 
I in the mountains (Adukpung spring, 
| Kiangan) was practically identical 
I with that (1300 X 10- 12 ) of the 
§ most active lowland spring (Sinabac 
| spring, Majayjay, Laguna), but con- 
l siderably less than that (2100 X 
I 10- 1 -) of the most active lowland 
g water (from a flowing well in Batan- 
I gas, Batangas). 

I In Nueva Vizcaya the radioactivity 
| encountered was uniformly low, so 
* that the present work has served to 
| emphasize the peculiarity noted last 
§ year, that is, a low radioactivity in 
c all waters except those in a smajl 
ci district in Ifugao. Since the waters 
g studied varied greatly, both in chem- 
ical quality and in the geological 
formations from which they were 
obtained, and since those highly 
radioactive showed no apparent pe- 
culiarities or marked differences 
from other, less active waters, the 
high activity of the Ifugao waters 
seems to be due to isolated local 
deposits of radioactive material. 



xii. a. 6 Heise: Radioactivity of Waters, Northern Luzon 305 

The general statements in the literature regarding radioac- 
tivity indicate that — 

The phenomenon is perhaps most common among waters of volcanic 
.origin, or at least among thermal springs. 12 

The present study seems to show that this generalization does 
not hold for Philippine waters, though the work is still too 
preliminary in character to justify positive statements. 

u Clarke, F. W., loc. cit., 215. 



ILLUSTRATIONS 

Plate I. Apparatus used in a field determination of radium emanation. 

TEXT FIGURES 

Fig. 1. Map of a part of northern Luzon. 

2. Generalized geologic section across northern Luzon, from Tagudin 
to the eastern coast. 

307 



Hi isi G. W. Radioai rrvrn of Waters. Northern Luzon.] [Phil. Journ. Sci., XII, A, Ni 




PLATE I. APPARATUS USED IN A FIELD DETERMINATION OF RADIUM EMANATION. 



THE CONSTANCY IN THE RADIOACTIVITY OF CERTAIN 
PHILIPPINE WATERS » 

By George W. Heise 

(From the Laboratory of General, Inorganic, and Physical Chemistry, 

Bureau of Science, Manila) 

Philippine springs and artesian wells frequently show great 
variations in flow. Many of them are appreciably augmented 
during the rainy season, and some, situated near the coast, flow 
only at high tide. However, so far as noted, the chemical qua- 
lity of deep-seated ground waters is not subject normally to 
marked variations, in spite of great fluctuations in the quantity 
of water. 2 This observation is in harmony with experience in 
other countries." 

The statements regarding the variation in emanation content 
of natural waters are somewhat confusing. Thus Ramsay * 
reported an increased emanation content in certain springs 
during periods of wet weather and great flow, whereas Steichen ' 
observed an increase in the activity in certain Bombay hot 
springs during the dry season, in a period of greatly reduced 
flow. The latter investigator pointed out that local conditions 
might well account for the differences noted. 

In a previous paper c it was pointed out that two determina- 
tions of the radioactivity of Sibul Springs, one made during 
the dry season, the other during a period of frequent rains, 
indicated that, under certain conditions, the seasonal variations 
might be very slight. Recently it has been found possible to 
test this conclusion by further work both on Sibul Springs and 
on a flowing artesian well. Apparatus, method, and limits of 
accuracy were the some as those previously described. 7 The 
results obtained are shown in Table I. 

1 Received for publication August, 1917. 

'Heise, G. W., This Journal, Sec. A (1916), 11, 125-7. 

a Hintz, F., and Kaiser, E., Zeitschr. f. prakt. Geol. (1915), 23, 122-6, 
through Chem. Abst. (1916), 10, 1741. 

4 Ramsay, R., The variation of the emanation content of certain springs, 
Phil. Mag. (1915), 30, 815-818. 

6 Steichen, A., The variation of the hot springs at Tuwa, Phil. Mag. 
(1916), 31, 401-3. 

6 Wright, J. R., and Heise, G. W., The radioactivity of Philippine waters, 
This Journal Sec. A (1917), 12, 145. 

7 Wright and Heise, op. cit. Heise, G. W., The radioactivity of the waters 
of the mountainous region of northern Luzon, This Journal, Sec. A (1917), 
12, No. 6. 



310 The Philippine Journal of Science \m 

Table I. — Variation in radium emanation content. 



Sibul Springs 

Do : 

Do 

Do 

Do... 

Do 

Flowing well, Parafiaque, Rizal 
Do ._ 



Date of ex- 
amination. 


Radium 
emanation 

content 
(as g. RaX 

10-12.) 


April 10, 1916 
July 9, 1916 
April 22, 1917 

do 

June 9.1917 
July 1.1917 
July 15.1917 
June 19, 1916 
June 2,1916 


1284 
1293 
1300 
1280 
1330 
1270 
1280 
632 
640 



The. differences between determinations of the radioactivity 
of the same source at various times are well within the limits 
of experimental error in this class of work. Though the work 
on Sibul Springs was done in only a few months in the year, 
determinations were made under greatly varied conditions. Si- 
bul Springs has a very large flow throughout the year, but the 
quantity of water changes appreciably with the season. The 
readings in April, 1916, and April, 1917, were made after pro- 
longed seasons of dry weather, at times of minimum water flow. 
Though there was no sharply defined rainy season in 1916, the 
July reading was taken during a period of frequent, though 
not very heavy, rains, at a time of increased flow. There was 
much heavier rainfall in 1917 than during the corresponding 
period in 1916, so that the July, 1917, readings were taken 
during a time of heavy precipitation and great flow. It is, 
therefore, reasonable to suppose that the readings noted are 
representative of the radioactivity of Sibul Springs through- 
out the year. 

The determinations of the activity of the Parafiaque flowing 
well are too few in number to be conclusive; they are so con- 
cordant, however, that they may be regarded as corroboratory 
evidence. 

The data at hand clearly indicate, therefore, that the radio- 
activity of a deep-seated ground water may remain remarkably 
constant for long periods. 

As the waters at Sibul Springs have been highly regarded for 
a long time because of their supposed medicinal virtues, they 
have been frequently analyzed. It is of interest to point out 
that surprisingly little, if any, change has been noted in the 
chemical character of the water for a long period of time. In 



Heise: Constancy of Philippine Waters 



311 



Table II are shown an analysis published by Centeno s in 1890 
and a recent routine analysis made in the Bureau of Science. 

Table II. — Analyses of ivater from Sibid Springs. 



! Total solids 

I Silica (SiOz) 

j Bicarbonates (HCO3) 

' Sulphates (S0«) 

' Chloride (CD 

Calcium (Ca) 

Magnesium (Mg) 



Analyzed 
Centeno, by Bureau 
1890.B of Science. 
1915.b 



* Recalculated as parts per million and to same terms as those used in standard practice. 
'> Analyzed by F. Pefia, chemist, Bureau of Science. 

Considering the length of time that has elapsed between 
the two analyses and the fact that the waters at Sibul Springs 
were neglected for a year during that interval, better agreement 
could be hardly expected. 



8 Centeno, J., et al., Memoria descriptiva de 
Isla de Luzon. Madrid (1890), 39. 



manantiales, etc. de 



INDEX 



Abar. 190. 

Abdong cahoy, 180. 

Abuab, 169. 

AGCAOILI, FRANCISCO, see BRILL, HAR- 
VEY C, 861. 

Agelaea, 184. 

AGUILAR, R. H., A comparison of linseed 
oil and lumbang oils as paint vehicles, 235. 

Albihal, 171. 

Alcohol from discard molasses in the Philip- 
pine Islands, 267. 

Alegango, 190. 

Aleurites cordata, 236. 
fordii, 236. 
moluccana, 167, 235. 
trisperma, 167, 236. 

ALINCASTRE. CECILIO, see BRILL, HAR- 
VEY C, 127. 

Alstonia scholaris, 167. 

Anisoptera thurifera, 111. 

Api-api, 111. 

Apitong, 111. 

Argemone mexicana, 167. 

ARGUELLES, ANGEL S., see BROWN, 
WILLIAM H.. 221. 

Asclepidaceffi, 174. 

Aspergillus flavus, 67. 
niger, 66. 

Atangen, 177. 

Avicennia officinalis, 111. 

B 

Babebabe, 172. 

Bacauan, 111. 

Bagtican lauan, 230. 

Bailey, E. H./S., see Reviews (book). 

Balangcari, 190. 

Balitadhan, 172. 

Balobo, 224. 

Batingui, 169. 

Bayabayabasan, 180. 

Berberis aristata, 178. 

Bo-nor, 172. 

Brick and mortar, composition of, in the great 
wall of China, 257. 

BRILL, HARVEY C, The antineuritic prop- 
erties of the infusorial earth extract of the 
hydrolyzed extract of rice polishings, 199 ; 
A chemical investigation of the seeds of 
Pangium edule and of Hydnocarpus alcate, 
37 : The fermentation of Philippine cacao, 1 ; 
see also PARKER. HARRISON O., 87. 

BRILL, HARVEY C, and AGCAOILI, 
FRANCISCO, Some limitations of the 
Kjeldahl method, 261. 

BRILL, HARVEY C, and ALINCASTRE, 
CECILIO, The possible maximum vitamine 
content of some Philippine vegetables, 127. 



BRILL, HARVEY C, and PARKER, HAR- 
RISON O., The rancidity of Philippine co- 
conut oil, 95. 

BRILL, HARVEY C, PARKER, HARRISON 
O., and YATES, HARRY S., Copra and 
coconut oil, 55. 

BRILL, HARVEY C, and THURLOW, LEA- 
VITT W., Alcohol from discard molasses in 
the Philippine Islands, 267. 

BRILL, HARVEY C, and WELLS, ALBERT 
H-, The physiological active constituents of 
certain Philippine medicinal plants: II, 167. 

BRILL, HARVEY C, and WILLIAMS, RO- 
BERT R., The use of chaulmoogra oil as a 
specific for leprosy, 207. 

BROWN, WILLIAM H., and ARGUELLES, 
ANGEL S., The composition and moisture 
content of the soils in the types of vegeta- 
tion at different elevations on Mount Ma- 
quiling, 221. 

Bruguiera parviflora, 111. 

Bugkan, 177. 

Buyun, 169. 

C 

Cacao, the fermentation of Philippine, 1 ; 

properties of butter from Philippine, 11. 
Cacaocacaoan, 180. 
Caesalpinia bonducella, 167. 

sappan, 167. 
Calamantao, 171. 
Calcium sulphate, the effect of, on cement, 

138. 
Camagsa taquilis, 185. 
Camagsang baguing, 184. 
Camunin, 186. 
Cataban, 225. 
Catmon, 169. 
Caua, 87. 

Cayutanang baguing, 177. 
Celastracese, 169. 

Cement, the effect of calcium sulphate on, 1S8. 
Charcoal, analysis of bacauan, 121. 
Chaulmoogra oil as a specific for leprosy, 207. 
China, composition of brick and mortar in 

the Great Wall of, 257. 
Chloride of lime, the interaction of, with the 

normal constituents of natural waters and 

sewage, 17. 
Cnestis, 184. 

Coconut oil and copra, 55. 
Coconut oil, methods for the production of 

pure, 87 ; the rancidity of Philippine, 95. 
Coconut products, the study of copra and 

other, 49. 
Cogon, 223. 
Combretacea?, 172. 
Congoura, 188. 
Connaraceae, 184. 

313 



314 



Index 



Connarus, 184. 

Copra-drying methods in use, 76. 

Copra, coconut oil and, 55 ; the study of, and 

other coconut products, 49. 
COX, ALVIN J., The study of copra and 

other coconut products, 49. 
Cratoxylon celebicura, 225. 



Dail, 174. 

Dalinding, 169. 

Datura fastuosa, 167. 

Dauag, 177. 

Dauag manoc, 177. 

Dayandag, 169. 

Dillenia philippinensis, 169. 

Diplodiscus paniculatus, 224. 

Dipterocarp forest, 224. 

Dipterocarpus sp., 111. 

Distillation, destructive, of Philippine woods, 

111. 
Dudu dudu, 44. 

E 

Elatostema, 225. 
Entada scandens, 167. 
Erythroploeum densifiorum, 171. 

guineense, 171. 
Erythroxylon burraanicum, 167. 
Exostemma philippicum, 190. 

G 

Gang6, 184. 

Gata, 87. 

Oelsemium sempervirens, 194. 

Great Wall of China, composition of brick and 

mortar in the, 257. 
Guayacan, 186. 
Guicos guicos, 185. 
Guijo, 111. 
Guyong guyong, 225. 
Gynocardase, 42. 
Gynocardia odorata, S7, 208. 
Gynocardin, properties of, 39. 

H 

Hanmabao, 185. 

HEISE, GEORGE W.. The constancy in the 
radioactivity of certain Philippine waters, 
309; The crater lake of Taal Volcano, 247; 
The interaction of chloride of lime with the 
normal constituents of natural waters and 
sewage, 17 ; The radioactivity of the waters 
of the mountainous region of northern Lu- 
zon, 293; see also WRIGHT, J. R., 145. 
Hopea sp., 111. 
Huliganga, 190. 
Hydnocarpus alcalaj and Pangium edule, a 

chemical investigation of the seeds of, 37. 
Hydnocarpus alcalae, 209. 

anthelmintica, 219. 
venenata, 37, 208. 
wighttiana, 208. 
Hymenodictyon excelsum, 190. 
obovatum, 190. 



I 
Iluhan, 88. 
Imperata exaltata, 223. 



Infusorial earth extract, the antineuritic prop- 
erties of the, of the hydrolyzed extract of 
rice polishings, 199. 

Intsia bijuga. 111. 

Ipil, 111. 

J 

Jatropha curcas, 107. 

K 

Kjeldahl method, some limitations of the, 261. 

Kukui, 235. 

L 

Labao, 180. 

Lake, crater, of Taal Volcano, 247. 

Langarai, 111. , 

Lanitan, 169. 

Lauan, white, 111. 

Leguminosa;, 171. 

Lenamo. 184. 

Leprosy, the use of chaulmoogra oil as a 
specific for, 207. 

Linseed oil, a comparison of, and lumbang 
oils as paint vehicles, 235. 

Lophopetalum fimbriatum, 170. 
toxicum, 169. 

Lubilubi, 180. 

Lumbang banucalag, 236. 

Lumbang bato, 235. 

Lumbang oils, a comparison of linseed oil and. 
as paint vehicles, 235. 

Lunas, 180. 

Lunas bondoc, 180. 

Lunas na puti, 180. 

Lunasia amara, 180. 

costulata, 181. 

Luzon, the radioactivity of the waters of 
the mountainous region of northern, 293. 

M 

Mag-talisay, 190. 

Malabunao, 171. 

Malacacao, ISO. 

Malauranada, 184. 

Malaligas na babrfe, 180. 

Malasanqui, 180. 

Malatabaco hibao, 190. 

Malatabigui, 171. 

Malatuba, 225. 

Maquiling, Mount, the composition and mois- 
ture content of the soils in the types of 
vegetation at different elevations on, 221. 

Mavindato, 185. 

Medicinal plants, the physiological active con- 
stituents of certain Philippine, 167. 

Microorganisms and their effect on copra and 
coconut oil, 63. 

Molasses, alcohol from discard, in the Phil- 
ippine Islands, 267. 

Mold, black, 66. 
brown, 67. 
green, 69. 
white, 64. 

X 

Narra, 111. 
Neolitsia villosa, 225. 
Jjgiricngic, 171. 
, Niogniogan. 172. 



Index 



315 



Oreocnide trinorvis, 



Paetan, 180. 

Paint vehicles, a comparison of linseed oil and 

lumbanB oils as. 235. 
Pait. 180. 
Paitan, 180. 
Pait-pait, 180. 
F'a'lo santo, 184. 
Palosapis, 111. 
Pangium edule. 209. 
Pangium edule and Hydnocarpus alcahe, a 

chemical investigation of the seeds of. 37. 
Parashorea plicata, 224. 
Parastaca plicata, 230. 

PARKER, HARRISON O.. and BRILL, HAR- 
VEY C, Methods for the production of pure 

coconut oil, 87; see also BRILL, HARVEY 

C. 55, 95. 
Pasuca, 174. 

Penicillium glaucum, C9. 
Pentacme contorta, 111. 
PEftA, F., review of Bailey's A Laboratory 

Guide to the Study of Qualitative Analysis, 

47. 
Philippine cacao, the fermentation of, 1. 
Philippine coconut oil, the rancidity of, 95. 
Philippine medicinal plants, the physiological 

active constituents of certain, 167. 
Philippine vegetables, the possible maximum 

vitamine content of some, 127. 
Philippine waters, the radioactivity of, 145. 
Philippine woods, destructive distillation of. 

111. 
Pine, Benguet, 111. 
Pinus insularis. 111. 
Pinonea, 172. 
Plants, Philippine medicinal, the physiological 

active constituents of certain, 167. 
Poonac, 88. 
Pterocarpus sp., 111. 
Puti-i-babae, 169. 
Puti-i-Ialaque, 169. 

Q 

Quercus solariana, 225. 
Quisqualis indica, 172. 



Rabelaisia, 180. 

Radioactivity of Philippine waters, 145 ; of 
the waters of the mountainous region of 
northern Luzon, 293 ; the constancy in the, 
of certain Philippine waters, 309. 

Radix indica lopeziana, 177. 

REVIEW: 
Baily, E. H. S., A Laboratory Guide to the 
Study of Qualitative Analysis, 47. 

REYES, F. D., see WITT, J. C, 133. 

Rhizophora sp.. 111. 

Rhizopus sp., 64. 

Rice polishinga, the antineuritic properties of 
the infusorial earth extract of the hydrolyzed 
extract of, 199. 
151034 4 



Rourea erccta. 184. 

heterophylla, 1S5. 

volubilis, 184. 
Rubiacese, 190. 
Rutacea;, 177. 



Saccharomyces, 273. 

Saccharum spontaneum, 223. 

Saguit, 180. 

Salsal, 171. 

Saltiqui, 180. 

Santiqui, 180. 

Saruncad, 174. 

Sarungcar, 174. 

Sauraunia barnsii, 225. 

Sayoncal, 174. 

Seeds, a chemical investigation of the, of 
Pangium edule and of Hydnocarpus alcalse, 
37. 

Selaginella, 225. 

Sewage and natural waters, the interaction of 
chloride of lime with the normal con- 
stituents of. 17. 

Shorea guiso. 111. 

polysperma, 111. 

Soils, the composition and moisture content of 
the, in the types of vegetation at different 
elevations on Mount Maquiling, 221. 

Strobilanthus plurifomis, 225. 

Strombosia philippinensis, 169. 

Sudcad, 169. 

Sulphur dioxide, preparing copra by use of, 
80. 

T 

] Taal Volcano, the crater lake of, 247. 
| Tagarao. 172. 

Talahib, 223. 
[ Tal-lolang, 172. 

Talolon, 172. 

Talulong, 172. 

Tamauyan, 169. 

Tangolo, 172. 

Tangolong, 172. 

Tanguili, 111. 

Tangulong, 172. 

Tapahan method, 51. 

Taraktogenos, 37. 

kurzif, 208. 

Tars, 118. 

Tartaro, 172. 

Tartaraoc, 172. 

Tauiigon, 172. 

THURLOW, LEAVITT W., see BRILL, 
HARVEY C., 267. 

Tinospora crispa, 167. 

Toddalia aculeata, 178. 
asiatica, 177. 

Tubo-bato, 190. 

Tylophora asthmatica, 176. 
brevipes, 174. 

u 

Ungali na mapula magtabig, 1S5. 






316 



Index 



Vegetables, Philippine, the possible maximum 

vitamine content of some, 127. 
Vitamine content, the possible maximum, of 

some Philippine vegetables, 127. 

W 

Water, absorption of, by copra, 7:>. 

Waters, the interaction of chloride of lime 
with the nisimal constituents of natural and 
sewage, 17 ; the constancy in the radioac- 
tivity of certain Philippine, 309 ; the radio- 
activity of Philippine, 145; the radioactivity 
of the, of the mountainous region of 
northern Luzon, 293. 

WELLS, A. H., Destructive distillation of 



Philippine woods, 111 ; see also BRILL, 
HAEVEY C, 1G7. 

WILLIAMS, ROBERT R., see BRILL, HAR- 
VEY C, 207. 

WITT, J. C, Composition of brick and mor- 
tar in the Great WaU of China, 267. 

WITT, J. C, and REYES, F. D., The effect 
of calcium sulphate on cement, 133. 

Woods, destructive distillation of Philippine, 
111. 

WRIGHT, J. R., and HEISE, GEORGE W., 
The radioactivity of Philippine waters, 145. 



Yacal, 111. 

YATES, HARRY S. 
C, 55. 



see BRILL, HARVEY 






PUBLICATIONS FOE SALE BY THE BTJEEAU OF SCIENCE, 
MANILA, PHILIPPINE ISLANDS— Continued 



BOT.\ v 
A FLORA OF MANILA 
By Elms* D. MHwru, 

Order No. 419. Paper, 490 pages. $2.50, 
postpaid. 
Prsotlcally a complete flora of the cul- 
tivated areas in the Philippines. Descrip- 
tions, with keys, of over 1,000 speoles, 590 
Oenera, and 136 families, with native names, 
glossary of teohnioal terms, eto. 



PHILIPPINE DIPTEROCARP FORESTB 
By William H. Brown and Donald M. 

Maithbws 

Order No. 432. Paper, 150 pases, 1 map, 
13 plates, and 12 diagrams, $1.00, 
postpaid. 
In Philippine Olpterooarp Forests the 
authors present a very comprehensive discus- 
sion of the growth and development of dip- 
terocarp trees and of the other elements of 
lowland Philippine forests. 



INDO-MALAYAN WOODS 
By Fred W. Foxwobthy 

Ordar No. 411. Paper, 182 pages, 9 

plates, $0.50, postpaid. 
In Indo-Malayan Woods, Doctor Fox- 
worthy has brought together a large amount 
of accurate Information concerning trees 
yielding woods of economic value. 



ZOOLOGY— Cont limed 
A MANUAL OF PHILIPPINE BIRDS 

By Richard C. McGregor 



A Manual of Philippine Birds oontalns 
In oompact form descriptions of all the 
known speoles of Philippine birds. The usual 
keys and diagnoses of orders, families, and 
genera help the novioe in identification. 



By David Starr Jordan and Robert Earl 
Richardson 



Paper, 7S pages, $0.75, 



This list will be found a oonvenlent guide 
to the synonymy of Philippine ichthyology. 
The nomenclature is thoroughly revised, and 
the distribution of each species within the 
Philippine Islands is given. 



By W. Schultzb 



LIST OF MAMMALS OF THE 
PHILIPPINE ISLANDS, EXCLU- 
SIVE OF THE CETACEA 

By Ned Hollister 

ler No. 418. Paper, 64 pages, $0.50, 



The distribution of each speoles Is given, 
and the original descriptions are oited. 



Order No. 436. Paper, 198 pages, $1.00, 
postpaid. 

This oatalogue Inoludes the names of all 
species of Coleoptera that have been recorded 
from a definite locality in the Philippine 
Islands. References to original descriptions 
and other important notes are given. The 
economlo appendix includes oomment on 
those species of beetles which are known to 
be Injurious or beneficial to man. 



PRICES ARE IN UNITED STATES CURRENCY 

Orders for these publications may be sent to the BUSINESS MANAGER., 
PHILIPPINE JOURNAL OF SCIENCE, BUREAU OP SCIENCE, MANILA, P. I., 
or to any of the agents listed below. Please give order number. 



The Macmillan Company, 64—66 Fifth Avenue, New York, TT. S. A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, "W. C, England. 

Martinus Nijhoff, Lange Voorhout 9, The Hague, Holland. 

Mayer & Miiller, Prinz Louis Ferdinandstrasse 2, Berlin X. W., Germany. 

Kelly & Walsh, Ltd., 32 Raffles Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 19 Baillle Street, Colombo, Ceylon. 

Tliacker, Spink & Co., P. O. Box 54. Calcutta. India. 



CONTENT 

BRILL, HARVEY C, and THURLOW, LEAVITT W. Alcohol 

from discard molasses in the Philippine Islands. 267 

HE1SE, GEORGE W. The radioactivity of the waters of the 
mountainous region of northern Luzon... 

HEISE, GEORGE W. The constancy in the radioactivity of cer- 
tain Philippine waters 





(.: s. 
The "Philippine Journal of Science" is Issued as follows: currency. 

Bectlon A. Chemical and Geological Sciences and the Industries.. $2.00 

Section B. Tropical Medicine 

Section C. Boi;. 

Section D. General Biology, Ethnology, and Anthropology (8< 

lion 1) began with Volume V) . 
Entire Journal, Volume II, III, IV, or V .5.00 

Entire Journal, beginning with Volume VI ... 7.00 

Single numbers (except of Volume I) .. 

Each xcctlon. is separately paged and ituli 
Authors receive 100 copies of their papers free. 
Volume I, 1006 (not divided into sections) and supplement, sold 

only with a complete file of section A, B, or C 10.00 

Supplement to Volume I (botany) 

Volume I (without supplement), sold only with a complete file of 

section A, B, or C. 
Siugle numbers of Volume I 

Publications sent in exchange for the Philippine Journal of Science 
should be addressed: Library, Bureau of Science, Manila, P. I. 

Subscriptions may be sent to the Business Manager, Philippine Jour- 
nal of Science, Bureau of Science, Manila, P. I., or to any of the agents 
listed below: 

AGENTS 

The Macmillan Company, 04-66 Fifth Avenue, New York City, V. S. A. 

Wm. Wesley & Son, 28 Essex Street, Strand, London, W. C, England. 

Martin us Nljhoff, Lange Voorhout 0, The Hague, Holland 

Mayer & Miill.r. Prlnz Louis Ferdlnandstrasse 2, Berlin, N. W., Germany. 

Kelly & Walsh, Limited, 32 Rallies Place, Singapore, Straits Settlements. 

A. M. & J. Ferguson, 18 Bail lie Street, Colombo, Ceylon. 

Thackcr, Spink & Co., P. O- Box 54, Calcutta, India. 



vt the pott office